Broadband Wireless Mobile: 3G and Beyond [1 ed.] 0471486612, 9780471486619

Citation preview

BROADBAND WIRELESS MOBILE 3G and Beyond

Edited by

Willie W. Lu SIEMENS, USA

JOHN WILEY & SONS, LTD

Copyright q 2002

John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ, England Telephone (+44) 1243 779777 Email (for orders and customer service enquiries): cs [email protected] Visit our Home Page on www.wileyeurope.com or www.wiley.com

All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except under the terms of the Copyright, Designs and Patents Act 1988 or under the terms of a licence issued by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London W1T 4LP, UK, without the permission in writing of the Publisher. Requests to the Publisher should be addressed to the Permissions Department, John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ, England, or emailed to [email protected], or faxed to (+44) 1243 770571. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the Publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Other Wiley Editorial Offices John Wiley & Sons Inc., 111 River Street, Hoboken, NJ 07030, USA Jossey Bass, 989 Market Street, San Francisco, CA 94103 1741, USA Wiley VCH Verlag GmbH, Boschstr. 12, D 69469 Weinheim, Germany John Wiley & Sons Australia Ltd, 33 Park Road, Milton, Queensland 4064, Australia John Wiley & Sons (Asia) Pte Ltd, 2 Clementi Loop 02 01, Jin Xing Distripark, Singapore 129809 John Wiley & Sons Canada Ltd, 22 Worcester Road, Etobicoke, Ontario, Canada M9W 1L1 British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library ISBN 0 471 48661 2 Typeset in 10/12pt Times by Deerpark Publishing Services Ltd, Shannon, Ireland. Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire. This book is printed on acid free paper responsibly manufactured from sustainable forestry in which at least two trees are planted for each one used for paper production.

Contents Preface List of Contributors

ix xiii

1 Summary and Introduction

1

1.1 1.2 1.3 1.4 1.5

1 3 4 5 8

Introduction Network Architecture Protocol Stack Compact Open Core Conclusions

2 UMTS Air Interface

11

2.1 Introduction 2.1.1 3GPP 2.1.2 3GPP2 2.2 UMTS Air Interface 2.2.1 Layer 1 2.2.2 Layer 2 2.2.3 Layer 3 2.3 CDMA2000 Air Interface 2.3.1 Layer 1 2.3.2 Layer 2 2.3.3 Layer 3 2.4 Compatibility Issues 2.4.1 3GPP 3G 2.4.2 3G 2G 2.5 Enhancing 3G Capabilities 2.5.1 Adaptive Antennas 2.5.2 Space Time Transmission Diversity 2.5.3 Turbo Coding 2.5.4 Multiuser Detection 2.6 Conclusions

11 13 14 14 15 48 62 70 71 76 85 90 91 93 95 98 113 121 128 132

3 Network Architecture

137

3.1 Introduction 3.1.1 Requirements for 3G Systems 3.1.2 International Standardisation Activities 3.1.3 General Aspects of 3G Systems 3.1.4 Chapter Outline

137 138 138 140 141

vi

Broadband Wireless Mobile: 3G and Beyond

3.2

Generic Network Model 3.2.1 Physical Model 3.2.2 Functional Model 3.3 Network Architecture 3.3.1 3GPP Release 99 3.3.2 3GPP Release 4 3.3.3 3GPP Release 5 3.3.4 An Overview of PS Domain Protocols 3.4 UMTS Terrestrial Radio Access Network 3.4.1 UTRAN Architecture 3.4.2 UTRAN Functions 3.4.3 Control and User Plane Separation in UTRAN 3.4.4 UE UTRAN Association 3.4.5 The Uu Interface 3.4.6 The Iu Interface 3.4.7 Key Features of Iu Interface 3.4.8 Protocol Architecture across Iu 3.4.9 Signalling Procedures across Iu 3.4.10 Iur Interface 3.4.11 Iub Interface 3.4.12 Establishment of Data Bearers in UTRAN 3.5 Network Access Security 3.5.1 Key Security Principles 3.5.2 Weaknesses in Second Generation Security 3.5.3 Security Objectives 3.5.4 Security Architecture 3.5.5 Network Access Security

142 142 144 146 147 156 160 166 169 169 174 176 177 178 178 179 181 187 198 205 209 215 216 217 217 218 220

4 Emerging Wireless Applications and Protocols

239

4.1 4.2

239 240 240 243 251 252 254 256 260 262 262 262 263 264 265 266 267 268

4.3

4.4 4.5

Introduction Wireless Application Protocol (WAP) 4.2.1 WAP Markets 4.2.2 WAP Architectures and Protocols 4.2.3 WAP Securities 4.2.4 WAP Interoperability 4.2.5 WAP and 3Gwireless 4.2.6 WAP Services and Applications 4.2.7 WAP System Solutions i Mode 4.3.1 What is i Mode? 4.3.2 i Mode Compatible HTML 4.3.3 i Mode Network Structure 4.3.4 Features of i Mode 4.3.5 i Mode Applications 4.3.6 i Mode Developing Strategy Other Wireless Mobile Internet Application Technologies Conclusions

5 Initiatives in 4G Mobile Design

271

5.1

271 271 271 277 277

Introduction Who Needs 4G? What is 4G? 5.1.1 Social Background and Future Trends 5.1.2 Trends in ITU R 5.1.3 Wireless Access Systems Related to 4G Mobile 5.1.4 Key Technologies

Contents

vii

5.2 Microwave Propagation 5.2.1 Microwave Mobile Propagation Characteristics in Urban Environments 5.2.2 Microwave Mobile Propagation Characteristics in Residential Environments 5.3 Adaptive Antennas 5.3.1 Introduction 5.3.2 Algorithms 5.3.3 Space time Equaliser Using Adaptive Antennas 5.3.4 Implementation of the Space time Equaliser 5.3.5 CDMA Adaptive Array Antennas 5.3.6 SDMA (Spatial Division Multiple Access) 5.3.7 Summary 5.4 Multiple Access Schemes 5.4.1 Comparison and Improvement Technology of Multiple Access Schemes 5.4.2 Multi carrier CDMA 5.4.3 Summary 5.5 CDMA Dynamic Cell Configuration 5.5.1 Teletraffic Load in Cellular Radio Systems 5.5.2 Teletraffic Management and Access Methods 5.5.3 Channel Assignment 5.5.4 Control Methods in CDMA Systems 5.5.5 Principle of Dynamic Cell Configuration 5.5.6 Evaluation of DCC 5.5.7 Characteristics in Up and Downlinks 5.5.8 Future Works 5.6 CDMA Cellular Packet Communications 5.6.1 Transmission Power Control for Connection less Services 5.6.2 Service Fairness in a System with Site Diversity Reception 5.6.3 Accommodation of Asymmetric Traffic 5.6.4 Summary 5.7 Network Architecture and Teletraffic Evaluation 5.7.1 Reducing Interruptions During Handoff 5.7.2 Reducing Forced Terminations During Handoff 5.7.3 Handover Control Appropriate for Multimedia Communications Using ATM and IP Technologies 5.7.4 A Mobile Communication Traffic Model 5.8 TCP over 4G 5.8.1 Transmission Rate Control 5.8.2 Transmission Power Control for CDMA Wireless Systems 5.8.3 Steady State Analysis for Combining of Transmission Power Control and Packet Transmission Rate Control 5.8.4 Performance Evaluation 5.8.5 Conclusions 5.9 Decoding Technique in Mobile Multimedia Communications

277 279 285 288 288 290 291 293 295 296 300 300 301 302 307 307 307 308 309 309 310 311 312 314 315 316 318 322 325 326 327 327

341 342 343 343

6 Conclusions

357

Index

361

331 332 336 339 340

Preface ‘Life is beautiful if powered by wireless technology’ Congratulations. By opening this page you have become part of a great conversation. When I first proposed the 3Gwireless’2000 Conference back in 1998, lots of issues were unclear about this new technology. Now we are meeting at 3Gwireless’2002 where we have tremendous progress on this technology and the 3Gwireless system has been already deployed in some countries. Wireless mobile communication normally has a ten-year evolution cycle, five years for research and development, and another five years for implementation and deployment. While 3Gwireless starts to deploy worldwide, research on next-generation mobile technology is already on track. The Delson Group’s World Wireless Congress and Fourth-Generation Mobile Forum (4GMF) opened the door for this very hot ‘beyond 3G’ research and development in this planet. When the first generation mobile phone came to the world, people greatly enjoyed its easy wireless communication. I still remember how much I loved the first mobile phone. It was like a brick, easy to stand-up and cost me a thousand dollars. In the early 90s, everyone talked about digital, and GSM (the second-generation mobile phone) came into life. The only difference for me to use GSM compared to the old one is more information displayed in the phone time, name and short message but it was still expensive for me in the early beginning. GSM is the product of our digitalised society and benefited a lot from it. GSM hand phone became the no. 1 fashion commodity in China. In many under-developed provinces in China, people worked hard every day just to get a GSM phone rather than buy anything else, and young people tried to save lunches to pay for the new phone though still over 20% of GSM phones have never been in service. The Internet created lots of wealth as well as noise. Now we really cannot survive without the Internet, and cutting off e-mail is like cutting off our oxygen. Information exchange has become one of the most important parts of life. The Internet pushed the mobile communication hard to the edge of technology because the voice-centric GSM does not have much capability to deliver value-added Internet traffic. Business is driven by the market and 3Gwireless was quickly rolled out into the market with huge investment and involvement from lots of engineers, researchers and marketing professionals. The objective of 3Gwireless is very clear: to extend the Internet traffic to the mobile terminal and unwire the Internet. Therefore, the major improvement of 3Gwireless is on

x

Broadband Wireless Mobile: 3G and Beyond

the new air interface called Radio Transmission Technology (RTT) to support high data rate transmission over the radio link. However, when people dream of the beautiful wireless future, one fundamental issue was brought to the table: where is the spectrum for all this activity? While Internet networking business cooled down, most investors changed their focus to wireless, especially broadband wireless technologies which greatly promoted the development of short-range wireless products. Wireless Local Area Network, wireless Personal Area Network and broadband wireless access, etc. have just flooded the world. They work in the licensed bands or unlicensed bands, and they consume a huge amount of spectrum. So again, do we have enough spectrum? From all over the world, every week, we have one wireless standard coming out whether it is international, national or just a local standard. By 2010, there may not be any standard at all because there are too many standards already in this small planet. So, how can we survive in this wireless storm? Convergence of wireless mobile and local wireless access is the only solution! The next generation mobile terminal (4G Mobile or called ‘Beyond 3G’) should be a multiband, multi-standard, multi-mode and multi-media personal communicator with an embedded converged broadband wireless system core. Whether you are in the office, home, airport or shopping centre, etc, the communicator will automatically connect to the broadband short range wireless access networks (i.e. Wireless LAN, etc) to provide highspeed wireless connections. If you are on the move and cannot reach these local wireless access systems, you will be automatically switched to the wireless mobile networks. If both wireless mobile and wireless access networks are available, the default connection mode is local wireless access. This converged wireless system has, at least, the following benefits: † greatly increases the spectrum utilisation with spectrum sharing and reuse † brings more bits to the wireless users as most broadband services are in the local wireless access domain † integrated numbering, billing and enhanced security † full use of wireless network resources † guaranteed seamless wireless Internet communications

This is the concept of the fourth-generation mobile communications (4G Mobile), and the proposed vision of the Fourth-Generation Mobile Forum (4GMF) by the Delson Group. From the user perspective this 4G Mobile vision can also be described as a multi-sphere level concept. In the first level the user connects all carried devices like a camera, phone, mirror glasses for images, watch, etc. in a PAN (Personal Area Network) by short range connectivity systems. The second level links the immediate environment like a TV, a PC, a refrigerator, etc. to the user. Level three ensures the direct communication to instant partners as other users and vehicles. Different radio access systems like terrestrial mobile systems, satellite systems and HAPS (High Altitude Platform Stations) are provided in level four for full area coverage. These levels are surrounded by the Cyber World (services and applications domain) in level five, where games, access to databases and the Internet, communication etc. are provided. Therefore, the different communication relations person-to-person and mainly machine-to-person and vice versa and machine-to-machine will determine mobile and wireless communications in the future.

Preface

xi

This vision from the user perspective is the driving force for seamless services and applications via different access systems (air interfaces) for future developments. Due to the future-dominating role of IP-based data traffic and applications, networks and systems have to be designed for economic packet data transfer. The fixed Internet penetration is growing in parallel to the mobile radio penetration. About 80% of fixed Internet users are also using mobile communications. Therefore, these users want to get the same services on wireless terminals. These services require a high degree of asymmetry between uplink and downlink especially for Internet type services with much higher expected capacity on the downlink. This 4G Mobile vision can be implemented by integration of these different evolving and emerging wireless technologies in a common flexible and expandable platform to provide a multiplicity of possibilities for current and future services and applications to users in a single terminal. The available, emerging and evolving radio transmission technologies have basically been designed in the classical vertical communication model that a system has to provide a limited set of services to users in an optimised manner. The 4G Mobile system will mainly be characterised by a horizontal communication model, where different air interfaces as cellular, cordless, WLAN type systems, short range connectivity and wired systems will be combined on a common platform to complement each other in an optimum way for different service requirements and radio environments. These wireless systems will be connected to a common, flexible and seamless converged core network. The mobility management will be part of a new Media Access System as interface between the core network and the particular wireless technology to connect a user via a single number for different systems to the network. This will correspond to a generalised access network. Global roaming for all wireless technologies is required. The interworking between these different systems in terms of horizontal and vertical handover and seamless services with service negotiation including mobility, security and QoS will be a key requirement, which will be handled in the newly developed Media Access Control System and the core network. Let’s back track to the purpose of this book. Why do we call it Broadband Wireless Mobile? People like the mobile phone much more because of its mobility than just because it is wireless. The mobile services continue to dominate the whole of wireless communications, and therefore, the fourth generation is still focused on the mobile business (called 4G Mobile). Meanwhile, W-LAN may become part of the mobile communications since each W-LAN base station can act as the mobile wireless router in terms of traffic management and access control. The world changes too fast. When I first talked on 3Gwireless in 1995, only a few people attended my seminar. Now everyone is talking about ‘‘3G’’, ‘‘Beyond 3G’’ and 4G Mobile. My invited speech in Stanford attracted nearly one thousand wireless professionals and my recent talk in China witnessed over ten thousand people unbelievable! So, what’s the conclusion? A New Wireless Storm is really coming! Thanks Delson Group for great efforts in organising the 3Gwireless conferences and 4G Mobile Forum to help promote the education, research and business of this emerging wireless technology worldwide. I hope you enjoy reading this book, and find it useful. Willie W. Lu Cupertino, California September, 2002

List of Contributors Chapters 1, 4 and 6 Willie L. Lu SIEMENS 1730 North First Street, MS 14303 San Jose, CA 95112 USA

Chapter 2 Matilde Sa´nchez Ferna´ndez, Antonio Caaman˜o-Ferna´ndez, Javier Ramos-Lo´pez and Ana Garcı´a-Armada All of Universidad Carlos III de Madrid Av. Universidad, 30 28911 Legane´s, Madrid SPAIN

Chapter 3 Apostolis Salkintzis Motorola Global Telecom Solutions Sector GPRS/UMTS System Design & Standards 32 Kifissias Ave., Athens GR-15125 GREECE

Chapter 5 Takehiko Kobayashi Department of Information and Communication Engineering Tokyo Denki University 2-2 Kanda-nishiki-cho, Chiyoda-ku Tokyo 101-8457 JAPAN

1 Summary and Introduction Broadband wireless communications have gained an increased interest during the last few years. This has been fuelled by a large demand on high frequency utilisation as well as a large number of users requiring simultaneous high data rate access for the applications of wireless mobile Internet and e-commerce. The convergence of wireless mobile and access will be the next storm in the wireless communications, which will use a new network architecture to deliver broadband services in a more generic configuration to wireless customers and supports value-added services and emerging interactive multimedia communications. Large bandwidth, guaranteed quality of service and ease of deployment coupled with recent great advancements in semiconductor technologies make this converged wireless system a very attractive solution for broadband service delivery.

1.1 Introduction ‘The future of wireless is not just wireless, it is a part of life’. When we trace back to the 1980s, everyone dreamed to have a nice mobile phone. But if we dream of the wireless picture in 2010, the story will be totally different. Why? Because by that time, the wireless infrastructure (not just for communications) will be totally multi-dimensional, whether in technologies (diversified and harmonised), applications (free mobile, local or global), or services (service/bandwidth on demand). Our wireless personal communicator or assistant (the size of a wallet or up to a book with enough bandwidth and memory) can help us enjoy our lives. Wireless becomes easy and affordable in the mass market, even when you are away from your office; your business will never be off-line. The global roaming and high-speed wireless link (thanks to the tremendous silicon advancements) will make our travels wonderful and feel at home. The key applications evolved from the advancement of broadband wireless, and the underlining technologies, including broadband wireless mobile (3Gwireless and 4Gmobile), broadband wireless access, broadband wireless networking, as well as broadband satellite solutions will surely dominate the whole communications market and therefore improve the business model in many aspects. Convergence of broadband wireless mobile and access will be the next storm in wireless communications. Fuelled by many emerging technologies including digital signal processing, software definable radio, intelligent antennas, superconductor devices as well as digital transceiver, the future wireless system will be much more compact with limited hardware entity and more flexible and intelligent software elements. Re-configurable and adaptive

2

Broadband Wireless Mobile: 3G and Beyond

terminals and base stations helps the system easily applied in the wireless mobile as well as wireless access applications. The compact hardware and very small portion of software (called Common Air Interface Basic Input-Output System or CAI-BIOS) will go the way as the computer industries did in the past. A compact multi-dimensional broadband wireless model will be adopted for the system design and implementation. Wireless Mobile Internet will be the key application of this converged broadband wireless system. The terminal will be very smart instead of dumb, compatible to mobile and access services including wireless multicasting as well as wireless trunking. This new wireless terminal will contain the following features: † 90% of traffic will be data; † security function will be enhanced, e.g. finger print chip embedded; † voice recognition function will be enhanced, the keypad or keyboard attachment will be an option, and wireless; † the terminal will support single and multiple users with various service options; † the terminal will be a fully adaptive software reconfigurable terminal.

As the wireless communications evolve to this convergence, the 4Gmobile (Fourth Generation Mobile Wireless Communications) will be an ideal mode to support high datarate connections from 2Mbps to 20Mbps based on the new spectrum requirement for IMT2000 as well as the co-existence of the current spectrum for broadband wireless access. This 4Gmobile system’s vision aims at: † providing a technological response to an accelerated growth in the demand for broadband wireless connectivity; † ensuring seamless services provisioning across a multitude of wireless systems and networks, from private to public, from indoor to wide area; † providing the optimum delivery of the user’s wanted service via the most appropriate network available; † coping with the expected growth in the Internet-based communications; † opening new spectrum frontiers.

Figure 1.1 shows the convergence of wireless mobile and access in one track and generates the 4Gmobile. In the following sections, we will discuss some detailed implementation issues including system architecture, reference model, protocol stack as well as system design.

Figure 1.1

Convergence of wireless mobile and access in one track.

3

Summary and Introduction

Figure 1.2

Network reference model.

1.2 Network Architecture The future wireless network should be an open platform supporting multi-carrier, multi-bandwidth and multi-standard air interfaces, and content-oriented bandwidth-on-demand (BoD) services will dominate throughout the whole network. In this way, the packetised transmission will go all the way from one wireless end terminal to another directly. Figure 1.2 shows this new wireless network architecture. The major benefits of this architecture are that the network design is simplified and the system cost is greatly reduced. The Base Transceiver System (BTS) is now a smart open platform with a basic broadband hardware pipe embedded with a CAI BIOS. Most functional modules of the system are software definable and re-configurable. The packet switching is distributed in the broadband packet backbone (or core network called Packet Division Multiplex PDM). The wireless call processing, as well as other console processing, is handled in this network. The Gateway (GW) acts as proxy for the Core Network and deals with any issues for the BTS, and the BTS is an open platform supporting various standards and optimised for full harmonisation and convergence. The terminal (Mobile Station MS) can be single or multi-users oriented supporting converged wireless applications. Figure 1.3 illustrates the unified wireless networks based on this architecture [1].

4

Broadband Wireless Mobile: 3G and Beyond

Figure 1.3 Unified wireless networks.

1.3 Protocol Stack Considering the signalling protocol in Figure 1.2, the client-server model is established between a wireless terminal and the core network. The BTS becomes the agent in both directions. This end-to-end direct signalling can ensure the wireless terminal to be smart and intelligent rather than the dumb one in the current wireless system. Figure 1.4(a) shows the system protocol stack. Different services (ATM, IP, STM, MPEG, etc.) can be supported through ‘Service Convergence Layer’. To guarantee the Wireless QoS (quality of service) and high spectrum utilisation, a Dynamic Bandwidth Allocation (DBA) scheme is required through the ‘MAC DBA Sublayer’ which improves the conventional layer architecture. DBA scheduler is the core of the MAC. To realise the dynamic resource allocation, this scheduler is essential for the broadband wireless link, which in general helps: † † † † †

support class of service offerings; provide diagnostic support for all network protocols; eliminate the need for traffic shaping and user parameter control; eliminate end-to-end packet and/or cell delay variation; increase spectrum utilisation.

The ‘Transmission Convergence Layer’ handles with various transmission modulations, error corrections, segmentations as well as interface mappings of wireless mobile and access in the physical layer. Figure 1.4(b) shows an example for the support of wireless access applications.

5

Summary and Introduction

Figure 1.4

(a) General protocol stack. (b) Protocol stack: example.

1.4 Compact Open Core As mentioned in the previous sections, this converged broadband wireless system will have the following features: † † † † †

multi-standards: 3Gwireless plus broadband wireless access high channel density with efficient resource utilisation dynamically scalable data-rates: from 32 kbps to 20 Mbps software definable and over-the-air programmable modules open core: various re-configurable kernels and common air interface BIOS

Figure 1.5 depicts this multi-dimensional and re-configurable radio [1], while Figure 1.6 shows its open interfaces. As wireless goes multi-dimensional, different standards come out everyday for different applications. However, if you look at their architectures in details, most of them are the same or almost the same. The ‘All IP’ layer will become the common platform; the service will be based on the secured Wireless Mobile Internet; the convergence will focus on the variable services demand as well as transmission technologies. From the implementation point of the view, in the future, the wireless software will take about 75% of the work, while the hardware only takes 25% for the construction of the open platform. Figure 1.7 shows this basic hardware structure. The ‘digital block’ will eventually be implemented in one system (system-on-chip). The ‘analogue block’ outputs as an open module subject to various CAI standards. With the

6

Broadband Wireless Mobile: 3G and Beyond

Figure 1.5

Multi dimensional and reconfigurable radio.

Figure 1.6 Compact broadband wireless

open interface.

Summary and Introduction

7

Figure 1.7 Open platform for broadband wireless mobile and access.

superconductivity technology advances, this block will probably become a separate ‘analogue header’ only. The broadband pipe throughout this hardware will be re-configurable and adaptive. The ‘CAI BIOS’ will be the software kernel to access and control the common hardware platform. Figure 1.8 lists the major functions embedded in this compact hardware implementation, where minimum software control is required. There are four key modules in the systems: air interfaces modules, baseband processing unit, digital broadband transceiver and smart antenna array. The detailed functional segments are required for the converged implementations of the proposed broadband wireless system.

Figure 1.8

Functional segments of the converged broadband wireless systems.

As an example, Figure 1.9 shows the open terminal architecture of this compact wireless system, where the ‘DSP core’, ‘CAI BIOS and Soft Radio API’ and ‘Main Processor (MPU)/ CPU’ are three most important entities. ‘RF/IF Subsystem’ is an independent unit configurable to different applications of wireless mobile or wireless access. ‘Digital Down-Converter

8

Broadband Wireless Mobile: 3G and Beyond

Figure 1.9

Compact wireless open terminal.

(DDC)’, ‘Digital Up-Converter (DUC’, ADC and DAC are components of the broadband digital transceiver system. ‘SIG’ handles various signalling protocol stacks, e.g. ‘All IP’ stack and ‘IP on Air’ stack. The proposed Fourth Generation Mobile Communications (4Gmobile) [1,2,4] will be an ideal model of this converged wireless mobile and access system. The 4Gmobile network and terminal reconfigurability (scalable and flexible self-organised) includes: † the adaptation of resource allocation to cope with varying traffic loads, channel conditions and service environments; † integration of fixed/mobile/broadcasting networks and rules for distribution and decentralised control of functional entities; † protocols that permit the network to adapt dynamically to changing channel conditions, that allow the coexistence of low and high-rate users, hands-off of high-data-rate users between base stations, congestion-control algorithms that are cognisant of and adjust to changing channel conditions etc.; † development of system concepts for digital broadband millimetre wave capable of delivering higher bit rates for the broadband wireless access applications.

Therefore, the 4Gmobile will provide seamless high data rate wireless service over an increasing number of integrated but however distinct and heterogeneous wireless mobile and access platforms and networks operating across multiple frequency bands. This service adapts to multiple wireless standards (and multi-mode terminal capabilities) and to delay sensitive or insensitive applications over radio channels of varying bandwidth, across multiple operators and service provider domains with user fully controlled service quality levels.

1.5 Conclusions In this chapter, a new compact multi-dimensional broadband wireless core is summarised which focuses on the convergence of wireless mobile and access technologies. As the wire-

Summary and Introduction

9

less industry will boom in the coming years, this converged wireless system will surely become the major player for the wireless mobile Internet services and applications. ITU (International Telecommunications Union) defined the future wireless systems beyond the 3Gwireless as 4Gmobile, which actually outlines the key features of our proposed convergence of broadband wireless mobile and access systems. The 4Gmobile will present a beautiful wireless life in 2010 when at that time, wireless will not just be a technology. In this book, the authors will focus on the Broadband Wireless Mobile issues. Chapter 2 will discuss the air interfaces and radio protocol of this 3Gwireless and beyond system; Chapter 3 will focus on the network architecture and reference models; Chapter 4 will update the progress of emerging wireless applications protocols and Chapter 5 will study the initiatives in 4G mobile communications. Finally, Chapter 6 will draw out conclusions.

Acknowledgments This chapter is based on my previous keynote speech on ‘Multi-Dimensional Broadband Wireless’ at Stanford University which attracted over 500 attendees from Silicon Valley. Many materials here are from 3Gwireless’2000 proceedings by Delson Group. Thanks should also be given to my colleagues in ITU WP8F, ITU JRG 8A-9B and IEEE 802.16 for their encouragement and supports. References and Further Reading [1] ‘4Gmobile Beyond IMT 2000’, ITU 8F/INFO/4 E DOC, Mar. 2000. [2] ‘The 4Gmobile Systems’, ITU 8F/INFO/1 E DOC, Feb. 2000. [3] J. Hu, ‘Applying IP over wmATM Technology to 3Gwireless’, IEEE Commun. Mag., vol. 37, no. 11, pp. 64 67, Nov. 1999. [4] Proceedings of IEEE 3Gwireless’2000, San Francisco, CA, June 14 16, 2000. [5] Special Issue on ‘3Gwireless and Beyond’, IEEE Pers. Commun. Mag., Oct. 2000. [6] W. W. Lu, et al., ‘System Reference Model and Protocol Stack for Broadband Wireless Access’, Proceedings of IEEE ICC’2000, pp. 560 564, New Orleans, LA, June 18 22, 2000. [7] White paper on Broadband Wireless Access systems, http://www.ieee802.org/16

2 UMTS Air Interface 2.1 Introduction Universal Personal Communications (UPC) establishes the new concept of personal mobility and personal numbering [1]. In the UPC environment the fixed association between terminal and user identification is removed. This establishes the basis for personal mobility. Personal communications involves providing an essentially transparent connection so that a practical range of services can be automatically provided to people on the move [2]. Both wired and wireless access can, and should be involved, with existing infrastructures forming the basis of service delivery to a person rather than to a place. The goal of third-generation mobile systems is to provide users with world-wide coverage via handsets that have the capability to seamlessly roam between multiple networks (fixed and mobile, cordless and cellular) across regions, which currently use different technologies. This wireless and wired mobility clearly complements UPC, giving the user total mobility across both types of networks. Third-generation mobile systems are one step beyond the digital cellular and cordless systems that are now into service. At the global level regarding the third-generation of mobile systems, in ITU (International Telecommunication Union) there is already an initiative, IMT-2000 [3 7], settling the framework of the future telecommunication infrastructure. IMT-2000 will provide wireless access to the global telecommunication infrastructure through both satellite and terrestrial systems, serving fixed and mobile users in public and private networks. It is being developed on the basis of the ‘family of systems’ concept designed to be able to connect different radio transmission modules to the same core network equipment. The radio interfaces defined are based on different access technologies. The access technology not only defines how the users access the system. The structures defined in Figure 2.1 apparently define radio interfaces that are supported by two different access technologies, but in fact, what is defined, is a hybrid access technique. This is the case, for example, for IMT-TC (IMT-Time Code) where the uplink and downlink and the different users are separated based on transmission on different time slots and on the spreading sequences. Thus W-CDMA (Wideband Code Division Multiple Access) is the access technique defined for three of the interfaces, IMT-DS (IMT-Direct Spread), IMT-MC (IMT-Multi Carrier) and IMT-TC. TDMA (Time Division Multiple Access) also supports IMT-TC, as has already been mentioned, IMT-SC (IMT-Single Carrier) and IMT-FT (IMT- Frequency Time). IMT-FT is also supported by a hybrid access technique based on FDMA (Frequency Division Multiple Access) and TDMA.

12

Broadband Wireless Mobile: 3G and Beyond

Figure 2.1 Radio interfaces defined for IMT 2000.

There is a convergent effort towards the standardization of third-generation mobile systems that support ITU proposals that will accelerate the IMT2000 standardization activities. Different standards development organizations (SDOs), ARIB (Japan), CWTS (China), TIA (USA), TTA (Korea), TTC (Japan), T1 (USA) and ETSI (Europe) are participating in the development of these new standards. For that, consortiums and partnerships are being created among them whose results are standards for the radio interfaces (Figure 2.2).

Figure 2.2

SDOs working for radio interfaces standardization.

UMTS Air Interface

13

This chapter focuses on the radio interfaces defined with W-CDMA access technology: IMT-DS, IMT-TC and IMT-MC. The radio interfaces defined with W-CDMA are being developed thanks to the creation of the 3G partnership as a multilateral collaboration among SDOs, aiming to facilitate the development of global technical specifications for 3G mobile systems as an evolution of the present mobile architectures: GSM and ANSI/TIA/EIA-41. The 3G partnership is divided into two projects, 3GPP supported by the SDOs that are actually involved in specification of GSM systems and its evolution, and 3GPP2 that will comprise the ANSI/TIA/EIA-41 network evolution, involving the corresponding SDOs. The evolution of both technologies does not imply a convergence in the same solution, given that, in addition to the partnership projects, work should be done in the direction of setting the interoperability procedures in order that the ‘family concept’, which is the basis for IMT-2000, is a reality. This includes not only considering a harmonization in the air interface specifications, but also a definition of the network interfaces (Figure 2.3). The present mobile technologies are evolving in such a way that 3GPP is working towards the definition of IMTDS and IMT-TC and 3GPP2 for IMT-MC. The radio interfaces are better known as UMTSFDD, UMTS-TDD and cdma2000.

Figure 2.3 Definition of architecture interfaces and interoperability.

2.1.1 3GPP In December 1998 five market-driven SDOs, ARIB (Japan), ETSI (Europe), T1 (USA), TTA (Korea) and TTC (Japan), agreed to launch the 3rd Generation Partnership Project (3GPP) with the posterior incorporation of CWTS (China) in May 1999. The aim of this project is to cooperate for the production of globally applicable technical specifications for a 3rd generation mobile system called Universal Mobile Telecommunication System (UMTS) [8], based on an innovative radio interface Universal Terrestrial Radio Access (UTRA) and evolution of the GSM core network. The technical specifications will be transposed into relevant standards by the participating SDOs using their established processes.

14

Broadband Wireless Mobile: 3G and Beyond

The technical specifications are focused on four main issues that comprise system and service aspects, terminals, the core network and the radio access network. In the first section of this chapter the UMTS radio access network is discussed in detail.

2.1.2 3GPP2 The partnership project 3GPP2 was launched in order to complement the evolution study of non-GSM systems. 3GPP2 is an effort by the International Committee of the American National Standards Institute’s (ANSI) to establish the evolution for ANSI/TIA/EIA-41 networks and their Radio Transmission Technologies (RTTs).

2.2 UMTS air interface The assumed UMTS architecture [9] defines three main functional entities: User Equipment (UE), UMTS Terrestrial Radio Access Network (UTRAN), and core network. The interfaces defined between the UE and the UTRAN is the radio interface (Uu) and the interface between the core network and the UTRAN is called Iu. The point to focus on are the interfaces: they have been clearly defined in order to set the interoperability of UEs of different providers with UTRANs from different telecommunication operators. In particular, this chapter focuses on the radio interface. The radio interface is characterized through its protocols [9,10] where it can be defined by two main groupings according to the final purpose service: the user plane protocols and the control plane protocols. The first carry user data through the access stratum and the second is responsible for controlling the connections between the UE and the network and the radio access bearers. A general protocol architecture splits the radio interface in three layers: a physical layer or Layer 1, the data link layer (Layer 2) and the network layer or Layer 3. This hierarchical stratification provides a complete vision of the radio interface, from both the functionality associated with each of the structured layer to the protocol flow between them. The purpose of the protocol stack is to set the services to organize the information to transmit through logical channels whose classifying parameter is the nature of the information they carry (i.e. control or traffic information) and map these logical channels into transport channels whose characteristic is how and with what characteristic the information within each logical channel is transmitted over the radio interface. This how and with whatcharacteristic means that for each transport channel there is associated one or more transport formats, each of them defined by the encoding, interleaving bit rate and mapping onto the physical channel. Each layer is characterized by the services provided to the higher layers or entities and the functions that supports them. Layer 1 [11],[12] supports all functions required for transmission of information on the physical medium offering information transfer services through the physical channels to higher layers. This includes preparing the transport channels to be sent through the physical medium, controlling hypothetical errors and measuring parameters related to the quality of the transfer service provided: frame error rate, signal to interference ratio, power measurements, etc. Layer 2 is subdivided into the medium access control (MAC) layer, the radio link control (RLC) layer, Packet Data Convergence Protocol (PDCP), and Broadcast /Multicast Control (BMC), each of them providing services to higher layers through their associated functions.

UMTS Air Interface

15

The MAC layer handles the transport channels, mapping the logical channels into transport formats and transferring to peer MAC entities protocol data units. The mapping is controlled by Layer 3 which is the one that determines the transport format set; consequently MAC layer should be able to select an appropriate transport format depending on the instantaneous source rate, the priority handling inside one UE and between UEs. If any multiplexing of protocol data units (PDUs) into transport blocks are needed it should be performed by the MAC layer since this functionality is not performed by Layer 1. RLC provides the transference of higher PDUs to the receiving entity and transfer of user data with quality of service settings. The transfer can be achieved in three modes: transparent, unacknowledged and acknowledged, performing for that segmentation, reassemble, concatenation and padding. PDCP provides transmission of higher PDUs in acknowledged, unacknowledged and transparent RLC mode, mapping the network protocol into an RLC entity. BMC provides a broadcast/multicast transmission service in the user plane for common user data; the broadcast service is supported by a scheduled transmission of cell broadcast messages handling the radio resources needed. Layer 3 is the interface between the access stratum of the radio interface and the non-access stratum. It is subdivided into the radio resource control (RRC) layer which interfaces with Layer 2 and an upper layer looking after providing access service to higher layers in the nonaccess level. Layer 3 provides three different types of services: general control services, notification services and dedicated control services. The RRC layer handles the control plane signalling providing measurements reports and radio resource assignments to peer RRC layers and controlling through feedback information RLC, MAC and physical layers. Figure 2.4 refers. Given this brief introduction to the radio interface, next we describe in detail each of the components.

2.2.1 Layer 1 The physical layer (L1) access scheme is based on Wideband Direct-Sequence Code Division Multiple Access (WCDMA) technology with two duplex modes: Frequency Division Duplex (FDD) and Time Division Duplex (TDD). In the FDD mode a physical channel is characterized by the code and frequency. Additionally, in the uplink a physical channel is defined by the relative phase, i.e. if the physical channel is mapped in the quadrature or phase component of the QPSK modulation. In TDD mode a physical channel is characterized by the code and time slot. The chip rate is 3.84 Mchips/s, but a fixed chip rate does not imply a fixed service bit rate. Different symbol rates can be specified for each physical channel applying different spreading factors (SF) to each symbol. Table 2.1 presents the possible symbol rates for both duplex modes. L1 offers data transport services to higher layers. It has two open interfaces: one with MAC layer through transport channels and the other with RRC layer that controls the configuration of the physical layer. Each transport channel is characterized by its transport format set. To each transport format a physical processing is applied to define the physical channel. A physical channel is defined by a carrier frequency, channelization code, scrambling code, time interval (starting and stopping transmission time) and relative phase (uniquely in uplink) and, additionally in TDD, the timeslot and burst type.

16

Broadband Wireless Mobile: 3G and Beyond

Figure 2.4 Protocol Stack for the Radio Interface. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

The physical layer operates with a hierarchical structure with a basic time interval of 10 ms called a radio frame which is subdivided into 15 slots. The interpretation of each slot is different for each of the duplex modes. In FDD within the whole frame is processed the basic Table 2.1 Service symbol rates

Min. Max.

FDD Uplink SF Symbol rate (ks/s)

Downlink SF Symbol rate (ks/s)

TDD Uplink SF Symbol rate (ks/s)

Downlink SF Symbol rate (ks/s)

4 256

4 512

1 16

1 16

960 15

960 7.5

3840 240

3840 240

UMTS Air Interface

17

unit provided by MAC, the transport block, or several transport blocks, and within each slot a similar substructure is applicable. In TDD each slot applies not only to determine the duplicity needed for uplink and downlink, but also to separate different users. This means that, unlike FDD, in TDD not only the code domain is being used to separate different users, but also the time slot. This would somehow justify the high rates provided in TDD mode, since in this mode the SF does not determine the net physical channel symbol rate, as each physical channel is not being transmitted in every slot and consequently its average transmission rate is decreased. The advantages of TDD vs. FDD are that TDD presents an optimal implementation of asymmetrical services, where the bandwidth requirements are tighter in the downlink than in the uplink. 2.2.1.1 FDD In FDD mode, uplink and downlink transmissions use separate frequency bands. Next, four main points are treated in order to describe the flow of information from high layers to L1 and how this information is finally mapped into a physical channel. Logical, transport and physical channels Logical channels, which are organized based on what type of information is transferred, are mapped into transport channels through the MAC layer. Given that the classification of transport channels is based on how the information is transferred, there is not a univocal matching between them. This is clearer for example observing that the attribute ‘common’ and ‘dedicate’ appears in both logical channels and transport channels, but sometimes a logical dedicated channel is mapped to a common transport channel and vice versa. The explanation for that is that a logical dedicated channel contains information for a particular user and can be carried on a common transport channel together with the information of other users. One step further is the mapping from transport channels to physical channels. Physical channels are not only determined by how the information is transmitted but also the type of information transmitted. The main grouping for transport channels is related to the exclusive use that is made of this channel; for that we consider dedicated channels (DCH) and common channels. The common channels are subdivided into Broadcast Channel (BCH), Forward Access Channel (FACH), Paging Channel (PCH), Random Access Channel (RACH), Common Packet Channel (CPCH) and Downlink Shared Channel (DSCH). Two logical channels are mapped into the DCH transport format, the Dedicated Traffic Channel (DTCH) and the Dedicated Control Channel (DCCH). The first is used for the transfer of user information and the second for control information. The DCH is an uplink or downlink channel that can be transmitted over the entire cell or over just part of the cell using beamforming. The DCH would set the transport format for the associated physical channels. Associated with the DCH, in the uplink there are Dedicated Physical Data Channel (DPDCH) and the Dedicated Physical Control Channel (DPCCH) corresponding to a mapping in the quadrature (Q) and phase (I) component, respectively, of the QPSK modulation. In the downlink, the physical channel is not characterized by the phase modulation in the QPSK so there is just one Downlink Dedicated Physical Channel (Downlink DPCH) where both control information and data information are multiplexed in time within the radio frame. The control information that is

18

Broadband Wireless Mobile: 3G and Beyond

transported in the dedicated physical channels is generated at L1. It consists of known pilots for channel estimation, transmit power control commands, and feedback information. RACH channel is an uplink transport channel that is always received from the entire cell. Its main characteristic is that it is characterized by an initial collision risk and open loop power control. This main characteristic will be useful for the transportation of different logical channels: Dedicated Control Channel (DCCH), Common Control Channel (CCCH) and Dedicated Transport Channel (DTCH). The RACH is mapped into the Physical RACH (PRACH) based on a slotted ALOHA with fast acquisition indication. The random access transmission can be started at defined time intervals and it consists of one or several preambles of 4096 chips and the message part of duration one or two radio frames. The structure of the message part is similar to the dedicated physical channel, since the type of information they transport is the same; what changes is the way the radio resource is accessed. As in the uplink dedicated physical channels, the data information is mapped in the I component and the control information inserted in the Q component by L1. CPCH carries information from DTCH and DCCH logical channels. It is an uplink transport channel associated with a downlink dedicated channel and characterized by the initial collision risk and inner loop power control. The physical channel with the transport format associated with the CPCH channel is the Physical CPCH (PCPCH). The transmission is based on CSMA-CD (Carrier Sense Multiple Access-Collision Detection). As in PRACH the transmission can be started at defined time intervals and the transmission structure is one or several preambles, a collision detection preamble, a power control preamble and N multiples of the frame duration containing the message. The logical channel that broadcasts system control information and cell specific information is the BCCH (Broadcast Control Information). One of the transport channel that carries this information is the BCH, which main characteristic is that it is always transmitted over the entire cell with a single transport format. The physical channel associated is the Primary Common Control Physical Channel (P-CCPCH). The transmission rate is fixed within this channel with a SF 256. The radio frame structure defined by frames of 10 ms divided in 15 slots is generated here by multiplexing in time the primary and secondary Synchronization Channel (SCH). The P-CCPCH is transmitted in every slot in the last nine symbols of each slot. The other transport channel that carries BCCH information is the FACH. But the transport characteristics of this channel, transmission over the entire cell or over part using beamforming, the possibility of fast rate change (each 10 ms) and slow power control, match the requirements for the transportation of many other logical channels: DCCH, CCCH, Common Traffic Channel (CTCH) and DTCH. FACH mapping into a physical channel is made to the Secondary Common Control Physical Channel (S-CCPCH) that is described next. A special control common logical channel is the Paging Control Channel (PCCH). It is a downlink channel that transmits paging information when the network does not know the cell location of the UE or the UE is in connect mode using sleep mode procedures. This channel is mapped to the PCH transport channel which at the physical level corresponds to S-CCPCH. Both transport channels have similar characteristics although the PCH is always transmitted over the entire cell and the FACH could be transmitted over just part of the cell. The SCCPCH has a frame structure based on the 10 ms frame with 15 slots. Within each slot data Transport Format Combination Indication (TFCI) field and pilot files are inserted and that is the key for the main difference between the P-CCPCH and the S-CCPCH; the latter supports

UMTS Air Interface

19

the different transport format necessary for the changing rate while the primary transport format is fixed. The DSCH is a downlink transport channel shared by different UEs. It is associated with one or several DCH and it carries information from DTCH and the DCCH. The DSCH could be transmitted over the entire cell or over part of it using beamforming. The associated physical channel is the PDSCH. The PDSCH is mapped in the 10 ms radio frame but in a radio frame different PDSCHs can be allocated using the channelization codes that are described in the corresponding section. It does not carry control information, but to indicate that there is information in the PDSCH for a UE either it uses the TFCI of the associated DPCH or higher layer signalling carried in the DPCH. The same way that the mapping between logical channels and transport channel is not biunivocal, there are some physical channels that do not have higher layer associated channels, i.e. there is not a transport or logical channel that is mapped into them. Those channels are: Common Pilot Channel (CPICH), Synchronization Channel (SCH), Acquisition Indicator Channel (AICH), Paging Indicator Channel (PICH), Access Preamble Acquisition Indication Channel (AP-AICH), CPCH Status Indicator Channel (CSICH) and Collision-Detection/Channel Assignment Indicator Channel (CD/CA-ICH). All of them are downlink channels that support functionalities associated explicitly with L1. The CPICH is used for transmission diversity to provide the UE with a channel where measurements of the channel state can be performed and later fed back to the base station. The SCH is used for cell search, synchronization purposes and defining the scrambling codes for downlink channels. The functionality is performed through two sub-channels. The primary synchronization channel caries a primary code that is repeated within the associated radio frame structure in every slot. This code is the same for every cell in the system. The secondary channel carries a secondary spread over the whole frame. This secondary code identifies the scrambling sequence for the downlink channels. The synchronization can be made on a frame basis using the secondary code and on a slot basis using the primary. The AICH, AP-AICH and CD/CA-ICH are used to support the uplink channels with random access, PRACH and PCPCH. CSICH is used to carry CPCH status information and is directly associated with the AP-AICH. PICH carries paging indicators for the SCCPCH. In Figure 2.5 all the mapping from logical channels to physical is detailed, specifying which channels are downlink, which uplink and which could be both. Multiplexing and coding of transport channels The multiplexing and channel coding functionality is a combination of error detection, error correction, rate matching, interleaving and multiplexing [30]. This functionality supports partly the transport services offered to MAC. Information arrives at the multiplexing and coding unit in transmission time intervals of 10 ms, 20 ms, 40 ms or 80 ms, in basic processing units called transport blocks. The process the information follows until is ready to be sent through the channel can be clearly differentiated into two main parts. First is the processing associated with each transport channel, i.e. setting the information with the corresponding transport format; here the information is coded, and arranged in order to provide the next step, multiplexation, with uniform information coming from every transport channel. The multi-

20

Broadband Wireless Mobile: 3G and Beyond

Figure 2.5

Mapping of logical channels into physical channels.

plexation is made over different transport channels. A code-composite transport channel (CCTCH) is generated and this would be mapped to one or different physical channels. The steps followed to set the transport format differ when referring to the uplink or downlink. There are three steps in the process that are common for both: Cyclic Redundancy Code (CRC), Transport Block Concatenation/Code Block Segmentation, and Channel Coding. In the uplink the following processes apply: radio frame equalization, first interleaving, radio frame segmentation and rate matching. In the downlink, after channel coding rate matching, first insertion of Discontinuous Transmission (DTX) indication bits, first interleaving and radio frame segmentation are performed. There are two main differences between uplink and downlink. The first is that, in the downlink, the transmission does not necessarily have to be continuous; discontinuous transmission bits can be inserted to determine in which time intervals there is no transmission. The second is that, in the uplink, the rate matching is made once first interleaving and radio frame segmentation is performed. In the multiplexing part the steps are transport channel multiplexing, second insertion of DTX indication bits (only in downlink), physical channel segmentation, second interleaving and physical channel mapping. In the CRC attachment a number of bits that are specified by higher layers are inserted in order to provide the transport block with a redundancy check. The number of bits inserted belongs to the set {0, 8, 12, 16, 24} and they are obtained through four different cyclic generator polynomials, respectively. Once the CRC is performed the serial concatenation of all the transport blocks to be processed might not give a total length adequate for the next step, coding. The channel coding defines a maximum size for the block to be coded (code block), depending on the coding that would be used. The possibilities are, convolutional coding with a maximum code block of 504 bits, turbo coding with a maximum code block of 5114 bits and no coding. If the

21

UMTS Air Interface

total number of bits after concatenation exceeds the corresponding maximum code block length, segmentation will be performed obtaining a number of code blocks of equal length. Which of the possible coding schemes is used depends on the transport channel that is being processed (Table 2.2). Table 2.2.2

Coding schemes for transport channels in FDD

Transport channel

BCH PCH RACH FACH CPCH DSCH DCH

Coding scheme Convolutional coding X X X X X X X

Turbo coding

X X X X

No coding

Coding rate CC TC

X X X X

1/2 1/2 1/2 1/2 1/3 1/2 1/3 1/2 1/3 1/2 1/3

1/3 1/3 1/3 1/3

Convolutional coding for both rate 1/2 and 1/3 has a memory order of 9. The generator polynomials are 561 (octal) and 753 (octal), respectively, for rate 1/2 and 557 (octal), 663 (octal) and 711 (octal) for 1/3. The turbo coding scheme is a Parallel Concatenated Convolutional Code with two constituent encoders. Each constituent encoder is a recursive systematic encoder with rate 1/2 and generator polynomials 15 and 13 (recursive bit) (octal). The systematic bit of the encoder for the interleaved information is punctured in order to obtain the 1/3 rate. The turbo interleaver is a design based on a rectangular matrix where intra-row and inter-row permutation is performed. Radio frame equalization is needed only in the uplink in order to set the number of bits at the input to a multiple of the number of radio frames (10 ms) in the transmission time interval. This operation is not necessary in the downlink since the output of the rate matching that is the function performed following the channel coding, is already a multiple of the number of radio frames. Rate matching repeats or punctures bits in order to match the physical channel bit rate. At this point in both uplink and downlink the number of bits in the radio frame is already fixed; considering that the radio frame transmission time interval is 10 ms, the number of bits in the radio frame is fixed by upper layer signalling with the spreading factor. First interleaving is performed while setting the transport format. It is a block interleaving with inter-column permutation. The number of columns and the permutation between them is fixed and depends on whether the transmission time interval is 10 ms, 20 ms, 40 ms or 80 ms. The radio frame segmentation is only necessary if the transmission time interval is longer than 10 ms, then the input sequence is divided and mapped into consecutive radio frames. Each 10 ms processed radio frame from each of the transport channels that will generate the CCTCH is delivered to the multiplexing unit. The number of transport channels that can be multiplexed depends on the nature on the transport channel. The same thing happens with the number of physical channels that can be generated from the CCTCH after the physical channel segmentation.

22

Broadband Wireless Mobile: 3G and Beyond

The second interleaving is made for each of the physical channels. It is a block interleaving with inter-column permutation that randomizes the bit position of the different transport channels. The last step is the physical channel mapping, at this point the control information needed by L1 for channel estimation TFI, etc. is inserted and spreading and modulation is made accordingly to each physical channel characteristic. Spreading and modulation FDD mode characterizes spreading and modulation distinguishing between uplink and downlink communications [31]. The modulation purpose is mapping the digital sequence (which in the particular case of the physical channels in the radio interface is the spread sequence) into signal waveforms that are appropriate for the channel. The modulation used is QPSK. Coherent demodulation of QPSK makes feasible transporting either the same source information doubling the transmission rate or mapping two different sources into the same modulated symbol. Spreading general purpose is to expand the original signal bandwidth in order to be able to transmit simultaneously information from different sources; in UMTS these sources are either different users, different channels associated with different users or base stations. The spreading process in UMTS is divided into two main parts. The first is the channelization operation where the bit sequence is spread with orthogonal sequences that preserves the orthogonality between sequences even when the SF are different. The second operation is scrambling where a scrambling code is applied to the spread signal. The channelization codes are Orthogonal Variable Spreading Factor (OVSF) codes that in the uplink preserve the orthogonality among the different user’s channels and in the downlink between the different channels in the base station. The number of codes available depend on SF (Figure 2.6). The generation of OVSF codes allows orthogonality among channels of different rates. The channelization code allocation in the uplink is fixed and depends on the SF applied to that particular channel. In the downlink the codes for the Primary Common Control Physical Channel (P-CCPCH) and the Common Pilot Channel (CPICH) are fixed; the codes for the rest of physical channels are assigned by UTRAN. Scrambling codes are different for the uplink and downlink, but in both are applied on a frame basis, i.e. every frame, that is compound of 38 400 chips, the scrambling sequence is repeated. This does not mean that the period of the scrambling sequence is 38 400 chips, the period of the scrambling sequence depends on its application, but every different sequence would be truncated and repeated in each frame. In the uplink there are two types of scrambling codes: long and sort. There are 2 24 possible long scrambling sequences. The different sequences are generated through different initialization of the 25 memory position shift registers that generate the sequence; the repetition period of each sequence is every 2 24 bits. For short sequences the number of possible sequences is the same but they are generated through 3 shift registers of 8 memory positions, consequently the repetition period is reduced to 256 chips. This means that within one frame of 38 400 chips, 150 periods of the code are present. The code allocation made by higher layers allows either short or long codes for mapping of dedicated channels and paging channels and different compositions of long codes for the rest of the channels.

23

UMTS Air Interface

Figure 2.6 OVSF codes generation trees.

In downlink the scrambling codes are generated by two shift registers of 18 memory 2 possible codes, but not all of them will be used. positions. This leads to n 0…2 18 From the whole set of codes a subset of 8192 codes will be used. This subset is classified in primary and secondary codes. The primary codes are the 512 codes where n is a multiple of 16. The secondary codes associated with a primary code are the 15 codes whose numbering follows the primary. This classification in primary and secondary codes makes sense since for each cell a single primary code is assigned; P-CCPCH and CPICH should always be transmitted using the primary code of the cell, being the secondary remaining codes used for the rest of the downlink physical channels. How the QPSK symbols are generated and from what information sources differs with the type of logical channel being mapped and if we are working in the uplink or downlink. In the uplink, for example, different data channels and a control associated channel can be mapped into just one physical channel by means of the channelization sequences that would separate

24

Broadband Wireless Mobile: 3G and Beyond

the different channels, and the orthogonality provided by the phase (I) and quadrature (Q) branches of the QPSK symbol. In the downlink every channel except the synchronization channel is serial to parallel converted into two branches, I and Q, and each of them spread by the same channelization code; the complex sequence generated is then multiplied by the complex scrambling sequence. Physical procedures Three physical procedures are described at this point [32]. These procedures define how L1 should work from the point of view of synchronization, power control and transmit diversity. From the point of view of synchronization, several issues should be achieved, but the critical ones are from the UE point of view. For that there exists a physical channel SCH that will exclusively be used by the UE for synchronization purposes. The first step is the cell search; for that is used a matched filter to the primary synchronization code that is common for all cells. At this point a slot synchronization is acquired too since this code is repeated every slot. The frame synchronization is made with the secondary codes at SCH identifying from all the possible codes which one is being used with a matched filter. Finally the PCCPCH can be detected by identifying the primary scrambling code through the CPICH that carries the channelization code and the primary scrambling code that is unique within the cell. In the synchronization process, upper layers are informed in both the uplink and downlink of the synchronization status. Power control is a physical procedure that handles the transmitted power in both uplink and downlink. This handling usually sets the control part power (control information inserted at L1 when physical channel mapping is implemented) while the relative power between control part and data part is defined by the corresponding transport format. There are two possibilities: open loop power control and inner loop power control. Open loop power control is the ability of the transmitter in the uplink to set its output power to a specific value. Inner loop power control is the ability of the transmitter to adjust its output power in accordance with one or more Transmit Power Control (TPC) commands received. It is used in both the downlink channel with information provided from the respective uplink channel and the uplink channel with information from the downlink channel. This counterpart implies that this power control method can be used where there are related channels in the uplink and downlink. TPC field is inserted within the frame structure in every physical channel that supports the inner loop in every slot; that implies that there is the possibility of updating the transmitted power in every slot. Particularly, there are two functioning ways; in the first the update is made every slot and in the second it is made every 3 slots for the base station and every 5 for the UE. Dedicated physical channels and PCHCH support inner loop power control and open loop is used for RACH power control. Transmit diversity should be supported in the downlink; there are two main possibilities that are given when there is information from the uplink of the channel state. In open loop mode there is no information from the uplink on the channel state and there are two possibilities: Time Switch Transmit Diversity (TSTD) that will be support for the SCH and Space Time Transmit Diversity (STTD) based on space time block coding. STTD should be supported for every channel in the downlink except for the SCH. These diversity techniques are addressed in greater detail in Section 2.5.

25

UMTS Air Interface

In close loop mode, the base station has information provided by the UE channel estimates seen from each antenna. This information is included in one of the control fields of the uplink dedicated control channel. UE computes this information through measurements in CPICH of the estimate channels seen from each antenna. This enhancing technique is addressed in more detail in Section 2.5. 2.2.1.2 TDD Logical, transport and physical channels Logical channels The MAC layer provides data transfer services on logical channels. A set of logical channel types is defined for different kinds of data transfer services as offered by MAC. Each logical channel type is defined by what type of information is transferred. The configuration of logical channel types is depicted in Figure 2.7.

Figure 2.7

Logical channel structure.

The following control channels are used for transfer of control plane information only: † † † † †

Broadcast Control Channel (BCCH) Paging Control Channel (PCCH) Common Control Channel (CCCH) Dedicated Control Channel (DCCH) Shared Channel Control Channel (SHCCH). The following traffic channels are used for the transfer of user plane information only:

† Dedicated Traffic Channel (DTCH) † Common Traffic Channel (CTCH).

Transport channels Transport channels are the services offered by Layer 1 to the higher layers. A transport channel is defined by how and with what characteristics data is transferred over the air interface. A general classification of transport channels is in two groups:

26

Broadband Wireless Mobile: 3G and Beyond

† dedicated channels, using inherent addressing of UE † common channels, using explicit addressing of UE if addressing is needed.

In TDD there are two types of dedicated transport channels, the Dedicated Channel (DCH) and the Fast Uplink Signalling Channel (FAUSCH). The DCH is an up- or downlink transport channel that is used to carry user or control information between the UTRAN and a UE. The FAUSCH is an uplink channel for signalling and is not yet included yet in the 1999 release of the specification. There are six types of control transport channels: BCH, FACH, PCH, RACH, USCH, DSCH. † The Broadcast Channel (BCH) is a downlink transport channel that is used to broadcast system and cell-specific information. † The Forward Access Channel (FACH) is a downlink transport channel that is used to carry control information to a mobile station when the system knows the location cell of the mobile station. The FACH may also carry short user packets. † The Paging Channel (PCH) is a downlink transport channel that is used to carry control information to a mobile station when the system does not know the location cell of the mobile station. † The Random Access Channel (RACH) is an up link transport channel that is used to carry control information from mobile station. The RACH may also carry short user packets. † The Uplink Shared Channel (USCH) is an uplink transport channel shared by several UEs carrying dedicated control or traffic data. † The Downlink Shared Channel (DSCH) is a downlink transport channel shared by several UEs carrying dedicated control or traffic data.

Indicators Indicators are a means of fast low-level signalling entities which are transmitted without using information blocks sent over transport channels. The meaning of indicators is implicit to the receiver. The indicator currently defined is the paging indicator. Physical channels All physical channels take hierarchical structure with respect to timeslots, radio frames and system frame numbering (SFN). Depending on the resource allocation, the configuration of radio frames or timeslots becomes different. All physical channels need guard symbols in every timeslot. The time slots are used in the sense of a TDMA component to separate different user signals in the time and the code domain. The physical channel signal format is presented in Figure 2.8. A physical channel in TDD is a burst, which is transmitted in a particular timeslot within allocated radio frames. The allocation can be continuous, i.e. the time slot in every frame is allocated to the physical channel or discontinuous, i.e. the time slot in a subset of all frames is allocated only. A burst is the combination of a data part, a midamble and a guard period. The duration of a burst is one time slot. Several bursts can be transmitted at the same time from one transmitter. In this case, the data part must use different OVSF channelization codes, but the same scrambling code. The midamble part has to use the same basic midamble code, but can use different midambles. The data part of the burst is spread with a combination of channelization code and scrambling code. The channelization code is a OVSF code, that can have a spreading factor of 1, 2,

UMTS Air Interface

27

Figure 2.8 Physical channel signal format. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

4, 8, or 16. The data rate of the physical channel is depending on the used spreading factor of the OVSF code used. The midamble part of the burst can contain two different types of midambles: a short one of length 256 chips, or a long one of 512 chips. The data rate of the physical channel depends on the midamble length used. So a physical channel is defined by frequency, timeslot, channelization code, burst type and radio frame allocation. The scrambling code and the basic midamble code are broadcast and may be constant within a cell. When a physical channel is established, a start frame is given. The physical channels can either be of infinite duration, or a duration for the allocation can be defined. The TDMA frame has a duration of 10 ms and is subdivided into 15 time slots (TS) of 2560 £ Tc duration each. A time slot corresponds to 2560 chips. Each 10 ms frame consists of 15 time slots, each allocated to either the uplink or the downlink. With such flexibility, the TDD mode can be adapted to different environments and deployment scenarios. In any configuration at least one time slot has to be allocated for the downlink and at least one time slot has to be allocated for the uplink. The DCH is mapped onto the dedicated physical channel. Downlink physical channels shall use SF 16. Multiple parallel physical channels can be used to support higher data rates. These parallel physical channels shall be transmitted using different channelization codes. Operation with a single code with spreading factor 1 is possible for the downlink physical channels. The spreading factors that may be used for uplink physical channels range from 16 down to 1. For each physical channel an individual minimum spreading factor SFmin is transmitted by means of the higher layers. There are two options that are indicated by UTRAN: † the UE shall use the spreading factor SFmin, independent of the current TFC; † the UE shall autonomously increase the spreading factor depending on the current TFC.

If the UE autonomously changes the SF, it shall always vary the channelization code along the lower branch of the allowed OVSF subtree. For multicode transmission, a UE shall use a maximum of two physical channels per timeslot simultaneously. These two parallel physical channels shall be transmitted using different channelization codes. Three types of bursts for dedicated physical channels are

28

Broadband Wireless Mobile: 3G and Beyond

defined. All of them consist of two data symbol fields, a midamble and a guard period, the lengths of which are different for the individual burst types. Thus, the number of data symbols in a burst depends on the SF and the burst type. The support of all three burst types is mandatory for the UE. The three different bursts are well suited for different applications. The BCH is mapped onto the Primary Common Control Physical Channel (P-CCPCH). The position (time slot/code) of the P-CCPCH is known from the Physical Synchronization Channel (PSCH). The P-CCPCH uses fixed spreading with a spreading factor SF 16. The P-CCPCH always uses channelization code cðk¼1Þ Q¼16 . PCH and FACH are mapped onto one or more secondary common control physical channels (S-CCPCH). In this way the capacity of PCH and FACH can be adapted to the different requirements. The S-CCPCH uses fixed spreading with a spreading factor SF 16. The RACH is mapped onto one or more uplink physical random access channels (PRACH). In such a way the capacity of RACH can be flexibly scaled depending on the operator’s need. This description of the physical properties of the PRACH also applies to bursts carrying other signalling or user traffic if they are scheduled on a time slot which is (partly) allocated to the RACH. The uplink PRACH uses either spreading factor SF 16 or SF 8. The set of admissible spreading codes for use on the PRACH and the associated spreading factors are broadcast on the BCH (within the RACH configuration parameters on the BCH). A midamble or training sequence is mapped in the burst associated with the PRACH. There exists a fixed association between the training sequence and the channelization code. In TDD mode code group of a cell can be derived from the synchronization channel. In order not to limit the uplink/downlink asymmetry, the SCH is mapped on one or two downlink slots per frame only. Due to mobile to mobile interference, it is mandatory for public TDD systems to keep synchronization between base stations. As a consequence of this, a capture effect concerning SCH can arise. The time offset toffset enables the system to overcome the capture effect. For Physical Uplink Shared Channel (PUSCH) the burst structure shall be used. Userspecific physical layer parameters like power control, timing advance or directive antenna settings are derived from the associated channel (FACH or DCH). PUSCH provides the possibility for transmission of TFCI in uplink. For Physical Downlink Shared Channel (PDSCH) the burst structure of DPCH shall be used. User-specific physical layer parameters like power control or directive antenna settings are derived from the associated channel (FACH or DCH). PDSCH provides the possibility for transmission of TFCI in downlink. The Paging Indicator Channel (PICH) is a physical channel used to carry the paging indicators. Beacon characteristics of physical channels For the purpose of measurements, physical channels at particular locations (time slot, code) shall have particular physical characteristics, called beacon characteristics. Physical channels with beacon characteristics are called beacon channels. The locations of the beacon channels are called beacon locations. The ensemble of beacon channels shall provide the beacon function, i.e. a reference power level at the beacon locations, regularly existing in each radio frame. Thus, beacon channels must be present in each radio frame.

UMTS Air Interface

29

Multiplexing, channel coding and interleaving Data stream from/to MAC and higher layers (transport block/transport block set) is encoded/ decoded to offer transport services over the radio transmission link [33]. The channel coding scheme is a combination of error detection, error correcting (including rate matching), and interleaving and transport channels mapping onto/splitting from physical channels. In the UTRA-TDD mode, the total number of basic physical channels (a certain time slot, one spreading code, on a certain carrier frequency) per frame is given by the maximum number of time slots which is 15 and the maximum number of CDMA codes per time slot. Figure 2.9 illustrates the overall concept of transport-channel coding and multiplexing. Data arrives at the coding/multiplexing unit in the form of transport block sets, once every transmission time interval. The transmission time interval is transport-channel specific from the set {10 ms, 20 ms, 40 ms, 80 ms}. The following coding/multiplexing steps can be identified: † † † † † † † † † †

add CRC to each transport block TrBk concatenation/code block segmentation channel coding radio frame size equalization interleaving (two steps) radio frame segmentation rate matching multiplexing of transport channels physical channel segmentation mapping to physical channels.

The coding/multiplexing steps for uplink and downlink are shown in Figure 2.9. Primarily, transport channels are multiplexed as described above, i.e. into one data stream mapped on one or several physical channels. However, an alternative way of multiplexing services is to use multiple CCTrCHs (Coded Composite Transport Channels), which corresponds to having several parallel multiplexing chains as in Figure 2.9 resulting in several data streams, each mapped to one or several physical channels. Error detection is provided on transport blocks through a Cyclic Redundancy Check (CRC). The size of the CRC is 24, 16, 12, 8 or 0 bits and it is signalled from higher layers what CRC size that should be used for each transport channel. All transport blocks in a TTI are serially concatenated. If the number of bits in a TTI is larger than the maximum size of a code block, then code block segmentation is performed after the concatenation of the transport blocks. The maximum size of the code blocks depends on whether convolutional, turbo coding or no coding is used for the TrCH. The bits input to the transport block concatenation are denoted by bim1 ; bim2 ; bim3 ; …; bimBi where i is the TrCH number, m is the transport block number, and Bi is the number of bits in each block (including CRC). The number of transport blocks on TrCH i is denoted by Mi. The bits after concatenation are denoted by xi1 ; xi2 ; xi3 ; …; xiXi , where i is the TrCH number and Xi MiBi. Segmentation of the bit sequence from transport block concatenation is performed if Xi . Z. The code blocks after segmentation are of the same size. The number of code blocks on TrCH i is denoted by Ci. If the number of bits input to the segmentation, Xi, is not a multiple

30

Broadband Wireless Mobile: 3G and Beyond

Figure 2.9 Transport channel multiplexing structure for uplink and downlink. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

31

UMTS Air Interface

of Ci, filler bits are added to the beginning of the first block. If turbo coding is selected and Xi , 40, filler bits are added to the beginning of the code block. The filler bits are transmitted and they are always set to 0. Code blocks are delivered to the channel coding block. They are denoted by oir1 ; oir2 ; oir3 ; …; oirKi , where i is the TrCH number, r is the code block number, and Ki is the number of bits in each code block. The number of code blocks on TrCH i is denoted by Ci. After encoding the bits are denoted by yir1 ; yir2 ; yir3 ; …; yirYi , where Yi is the number of encoded bits. The relation between oirk and yirk and between Ki and Yi is dependent on the channel coding scheme. The following channel coding schemes can be applied to transport channels: † convolutional coding † turbo coding † no coding.

The coding schemes applied are the same as for the FDD mode already described. Radio frame size equalization is padding the input bit sequence in order to ensure that the output can be segmented in Fi data segments of the same size. The input bit sequence to the radio frame size equalization is denoted by ci1 ; ci2 ; ci3 ; …; ciEi , where i is the TrCH number and Ei the number of bits. The output bit sequence is denoted by ti1 ; ti2 ; ti3 ; …; tiTi , where Ti is the number of bits. The output bit sequence is derived as follows: tik

cik

for k

ð1Þ

1; …; Ei

and tik

{0; 1}

for k

Ei 1 1; …; Ti ; if Ei , Ti

ð2Þ

where Ti

Fi £ Ni

ð3Þ

dEi =Fi e

ð4Þ

and Ni

is the number of bits per segment after size equalization.The first interleaving is a block interleaver with inter-column permutations. The input bit sequence to the block interleaver is denoted by xi;1 ; xi;2 ; xi;3 ; …; xi;Xi , where i is the TrCH number and Xi is the number of bits. Here Xi is guaranteed to be an integer multiple of the number of radio frames in the TTI. The output bit sequence from the block interleaver is derived as follows: † NUMBEREDSelect the number of columns C1 from Table 2.3 depending on the TTI. The columns are numbered 0,1,…C1 1 from left to right. † Determine the number of rows of the matrix, R1 defined as:

R1

Xi =C1

The rows of the matrix are numbered 0,1,…,R1

ð5Þ 1 from top to bottom.

† Write the input bit sequence into the R1 £ C1 matrix row by row starting with bit xi;1 in column 0 of row 0 and ending with bit xi;ðR1£C1Þ in column C1 1 of row R1 1:

32

Broadband Wireless Mobile: 3G and Beyond

Table 2.3 Inter column permutation patterns for first interleaving TTI (ms)

Number of columns C1

Inter column permutation patterns kP1C1(0), P1C1(1),…, P1C1(C1 1)l

10 20 40 80

1 2 4 8

k0l k0,1l k0,2,1,3l k0,4,2,6,1,5,3,7l

2 6 6 6 6 6 6 6 4

xi;1

xi;2

xi;3

xi;ðC111Þ

xi;ðC112Þ

xi;ðC113Þ

.. .

.. .

.. .

xi;ððR1

1Þ£C111Þ

xi;ððR1

1Þ£C112Þ

xi;ððR1

…

1Þ£C113Þ

xi;C1

3

7 … xi;ð2£C1Þ 7 7 7 7 .. 7 7 … . 5 … xi;ðR1£C1Þ

ð6Þ

† Perform
 the inter-column permutation for the matrix based on the pattern P1C1 j j[f0;1;…;C1 1g shown in Table 2.3, where P1C1(j) is the original column position of the jth permuted column. After permutation of the columns, the bits are denoted by yi,k:

2 6 6 6 6 6 6 6 4

3

yi;1

yi;ðR111Þ

yi;ð2£R111Þ

… yi;ððC1

1Þ£R111Þ

yi;2

yi;ðR112Þ

yi;ð2£R112Þ

… yi;ððC1

1Þ£R112Þ

.. .

.. .

.. .

…

.. .

yi;R1

yi;ð2£R1Þ

yi;ð3£R1Þ

…

yi;ðC1£R1Þ

7 7 7 7 7 7 7 5

ð7Þ

† Read the output bit sequence yi;1 ; yi;2 ; yi;3 ; …; yi;ðC1£R1Þ of the block interleaver column by column from the inter-column permuted R1 £ C1 matrix. Bit yi;1 corresponds to row 0 of column 0 and bit yi;ðR1£C1Þ corresponds to row R1 1 of column C1 1.

When the transmission time interval is longer than 10 ms, the input bit sequence is segmented and mapped onto consecutive Fi radio frames. Following radio frame size equalization the input bit sequence length is guaranteed to be an integer multiple of Fi. Rate matching means that bits on a TrCH are repeated or punctured. Higher layers assign a rate-matching attribute for each TrCH. This attribute is semi-static and can only be changed through higher layer signalling. The rate-matching attribute is used when the number of bits to be repeated or punctured is calculated. The number of bits on a TrCH can vary between different transmission time intervals. When the number of bits between different transmission time intervals is changed, bits are repeated to ensure that the total bit rate after TrCH multiplexing is identical to the total channel bit rate of the allocated physical channels. If no bits

UMTS Air Interface

33

are input to the rate matching for all TrCHs within a CCTrCH, the rate matching shall output no bits for all TrCHs within the CCTrCH. Puncturing can be used to minimize the required transmission capacity. The maximum amount of puncturing that can be applied is 1-PL; PL is signalled from higher layers. The possible values for Ndata depend on the number of physical channels Pmax, allocated to the respective CCTrCH, and on their characteristics (spreading factor, length of midamble and TFCI, usage of TPC and multiframe structure). Every 10 ms, one radio frame from each TrCH is delivered to the TrCH multiplexing. These radio frames are serially multiplexed into a coded composite transport channel (CCTrCH). When more than one PhCH is used, physical channel segmentation divides the bits among the different PhCHs. The second interleaving is a block interleaver and consists of bits input to a matrix with padding, the inter-column permutation for the matrix and bits output from the matrix with pruning. The second interleaving can be applied jointly to all data bits transmitted during one frame, or separately within each timeslot, on which the CCTrCH is mapped. The selection of the second interleaving scheme is controlled by higher layers. The bits after physical channel mapping are denoted by wp;1 ; wp;2 ; …; wp;Up , where p is the PhCH number and Up is the number of bits in one radio frame for the respective PhCH. The bits wp,k are mapped to the PhCHs so that the bits for each PhCH are transmitted over the air in ascending order with respect to k. Different transport channels can be encoded and multiplexed together into one Coded Composite Transport Channel (CCTrCH). There are two types of CCTrCH: † CCTrCH of dedicated type, corresponding to the result of coding and multiplexing of one or several DCH. † CCTrCH of common type, corresponding to the result of the coding and multiplexing of a common channel, i.e. RACH and USCH in the uplink and DSCH, BCH, FACH or PCH in the downlink, respectively.

Transmission of TFCI is possible for CCTrCH containing transport channels of: † † † †

dedicated type USCH type DSCH type FACH and/or PCH type. The following CCTrCH combinations in the uplink are allowed simultaneously:

† several CCTrCH of dedicated type † several CCTrCH of common type.

Allowed CCTrCH combinations on the downlink: † several CCTrCH of dedicated type † several CCTrCH of common type.

Transport format detection can be performed both with and without Transport Format Combination Indicator (TFCI). If a TFCI is transmitted, the receiver detects the transport format combination from the TFCI. When no TFCI is transmitted, so-called blind transport format detection may be used, i.e. the receiver side uses the possible transport format combi-

34

Broadband Wireless Mobile: 3G and Beyond

nations as a priori information. Blind transport format detection is optional both in the UE and the UTRAN. Therefore, for all CCTrCH, a TFCI shall be transmitted, including the possibility of a TFCI length zero, if only one TFC is defined. The Transport Format Combination Indicator (TFCI) informs the receiver of the transport format combination of the CCTrCHs. As soon as the TFCI is detected, the transport format combination, and hence the individual transport channels’ transport formats are known, and decoding of the transport channels can be performed. Encoding of the TFCI depends on its length, whether there are 6 10 bits of TFCI or if there are less than 6 bits. TFCI is encoded by the (32,10) sub-code of second order Reed Muller code. The code words of the (32,10) sub-code of second order Reed Muller code are linear combination of some among 10 basis sequences. The paging indicator Pq is an identifier to instruct the UE whether there is a paging message for the groups of mobiles that are associated with the PI, calculated by higher layers, and the associated paging indicator Pq. The length LPI of the paging indicator is LPI 2, LPI 4 or LPI 8 symbols. The TPC command is an identifier sent in uplink transmission only, to instruct NodeB whether Tx power has to be increased or decreased. The length of the TPC command is one symbol. Spreading and modulation In Table 2.4, a separation between the data modulation and the spreading modulation has been made [34]. Table 2.4 Basic modulation parameters Chip rate

Same as FDD chip rate: 3.84 Mchip/s

Low chip rate: 1.28 Mchip/s

Data modulation Spreading characteristics

QPSK Orthogonal; Q chips/symbol, where Q ¼ 2 p, 0 # p # 4

QPSK Orthogonal; Q chips/symbol, where Q ¼ 2 p, 0 # p # 4

The symbol duration TS depends on the spreading factor Q and the chip duration TC: Ts Q £ Tc, where Tc 1/chip rate. For any burst type the data modulation is performed to the bits from the output of the physical channel mapping procedure and always combines two consecutive binary bits to a complex valued data symbol. Each user burst has two data carrying parts, termed data blocks: ðk:iÞ d



ðk;iÞ
ðk;iÞ T ðd
ðk;iÞ 1 ; d 2 ; …; d NK Þ ;

i

1; 2; k

1; …; K

ð8Þ

Nk is the number of symbols per data field for the user k. Data block dðk;1Þ is transmitted before the midamble and data block dðk;2Þ after the midamble. Each of the Nk data symbols d
ðk;iÞ n ,i 1,2, k 1,…,K, n 1,…,Nk, of equation (1) has the symbol duration TsðkÞ Qk Tc as already given. are generated from two conseThe data modulation is QPSK, thus the data symbols d
ðk;iÞ n cutive data bits from the output of the physical channel mapping procedure:

35

UMTS Air Interface

bðk;iÞ l;n [ {0; 1};

l

1; 2; k

1; …; K; n

1; …; Nk ; i

1; 2

ð9Þ

using the mapping to complex symbols in Table 2.5. Table 2.5 Complex QPSK mapping Consecutive binary bit pattern

Complex symbol

ðk;iÞ bðk;jÞ l;n b2;n 00 01 10 11

d ðk;iÞ n 1j 11 1 j

The mapping corresponds to a QPSK modulation of the interleaved and encoded data bits bðk;iÞ l;n of equation (9). In case of PRACH burst type, the definitions in for burst types 1 and 2 apply with a modified number of symbols in the second data block. For the PRACH burst type, the number of symbols in the second data block dðk;2Þ is decreased by 96=QK symbols. of equation (1) is spread with a real For spreading, each complex valued data symbol d
ðk;iÞ n valued channelization code cðkÞ of length Qk [ {1; 2; 4; 8; 16}. The resulting sequence is then scrambled by a complex sequence n of length 16. 1,…,K, q 1,…,Qk, of the real valued channelization codes The elements cðkÞ q , k cðkÞ

ðkÞ ðkÞ ðcðkÞ 1 ; c2 ; …; cQk ;

k

1; …; K

ð10Þ

shall be taken from the set Vc

{1;

ð11Þ

1}

cðkÞ Qk

The are Orthogonal Variable Spreading Factor (OVSF) codes, mixing in the same timeslot channels with different spreading factors while preserving the orthogonality. As described above, the OVSF codes can be defined using the code tree of Figure 2.6. All codes within the code tree cannot be used simultaneously in a given timeslot. A code can be used in a timeslot if and only if no other code on the path from the specific code to the root of the tree or in the sub-tree below the specific code is used in this timeslot. This means that the number of available codes in a slot is not fixed but depends on the rate and spreading factor of each physical channel. The spreading factor goes up to QMAX 16. Qk is followed by a The spreading of data by a real valued channelization code cðkÞ of length
 n1 ; n2 ; :::; C n16 . The elements ni ; i cell-specific complex scrambling sequence n 1; :::; 16 of the complex valued scrambling codes shall be taken from the complex set: V
n

{1; j;

1;

j}

ð12Þ

In equation (12) the letter j denotes the imaginary unit. A complex scrambling code n is generated from the binary scrambling codes n ðn1 ; n2 ; …; n16 Þ of length 16 described in [34]. The relation between the elements n and n is given by

36

n
i

Broadband Wireless Mobile: 3G and Beyond

ðjÞi ·ni ;

ni [ {1; 1};

i

ð13Þ

1; …; 16

Hence, the elements ni of the complex scrambling code n are alternating real and imaginary. The length matching is obtained by concatenating QMAX/Qk spread words before the scrambling. The scheme is illustrated in Figure 2.10.

Figure 2.10 Spreading of data symbols.

The combination of the user-specific channelization and cell-specific scrambling codes can be seen as a user- and cell-specific spreading code sðkÞ ðsðkÞ p Þ with sðkÞ p

cðkÞ 11½ðp

1ÞbmodQk
·n
11½ðp 1ÞbmodQMAX
;

k

1; …; K; p

ð14Þ

1; …; Nk Qk

With the root raised cosine chip impulse filter Cr0(t) the transmitted signal belonging to the ðk;1Þ of equation (1) transmitted before the midamble is data block d
d
ðk;1Þ ðtÞ

Nk X

d
ðk;1Þ n

n¼1

Qk X

sðkÞ ðn

q¼1

1ÞQk 1q ·Cr0 ðt

ðq

1ÞTc

ðn

1ÞQk Tc Þ

ð15Þ

ðk;1Þ and for the data block d
of equation (1) transmitted after the midamble

d
ðk;2Þ ðtÞ

Nk X n¼1

d
ðk;2Þ n

Qk X q¼1

sðkÞ ðn

1ÞQk 1q ·Cr0 ðt

ðq

1ÞTc

ðn

1ÞQk Tc

Nk Qk Tc

Lm Tc Þ ð15Þ

where Lm is the number of midamble chips. The complex-valued chip sequence is QPSK modulated as shown in Figure 2.11. The pulse-shaping characteristics are a RRC filter with a roll-off factor of 0.22.

UMTS Air Interface

37

Figure 2.11 Modulation of complex valued chip sequences. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

Figure 2.12 illustrates the principle of combination of two different physical uplink channels within one timeslot. The DPCHs to be combined belonging to the same CCTrCH, did undergo spreading as described in sections before and are thus represented by complex-valued sequences. First, the amplitude of all DPCHs is adjusted according to UL open loop power control. Each DPCH is then separately weighted by a weight factor gi and combined using complex addition. After combination of physical channels the gain factor bj is applied, depending on the actual TFC. In case of a different CCTrCH, the principle shown in Figure 2.12 applies to each CCTrCH separately. The values of weight factors g i depend on the spreading factor SF of the corresponding DPCH.

Figure 2.12 Combination of different physical channels in uplink. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

Synchronization codes The primary synchronization code (PSC), Cp, is constructed as a so-called generalized hierarchical Golay sequence. The PSC is furthermore chosen to have good a-periodic auto correlation properties.

38

Broadband Wireless Mobile: 3G and Beyond

Three secondary SCH codes are QPSK modulated and transmitted in parallel with the primary synchronization code. The QPSK modulation carries the following information: † the code group that the base station belongs to (32 code groups: 5 bits; cases 1, 2); † the position of the frame within an interleaving period of 20 ms (2 frames:1 bit, cases 1, 2); † the position of the SCH slot(s) within the frame (2 SCH slots:1 bit, case 2).

The modulated secondary SCH codes are also constructed such that their cyclic shifts are unique, i.e. a non-zero cyclic shift of less than 2 (case 1) and 4 (case 2) of any of the sequences is not equivalent to some cyclic shift of any other of the sequences. Also, a non-zero cyclic shift less than 2 (case 1) and 4 (case 2) of any of the sequences is not equivalent to itself with any other cyclic shift less than 8. The secondary synchronization codes are partitioned into two code sets for case 1 and four code sets for case 2. Physical layer procedures (TDD) [35] Transmitter power control Power control is applied for the TDD mode to limit the interference level within the system, thus reducing the intercell interference level and reducing the power consumption in the UE. All codes within one timeslot allocated to the same CCTrCH use the same transmission power, in cases where they have the same spreading factor. By means of higher layer signalling, the Maximum Allowed UL TX power for uplink may be set to a value lower than what the terminal power class is capable of. The total transmit power shall not exceed the allowed maximum. If this were the case, then the transmit power of all uplink physical channels in a timeslot is reduced by the same amount in dB (Table 2.6). Table 2.6 Transmit power control characteristics

Power control rate Step size Remarks

Uplink

Downlink

Variable: 1 7 slots delay (2 slot SCH); 1 14 slots delay (1 slot SCH)

Variable, with rate depending on the slot allocation. 1, 2, 3 dB Within one timeslot the powers of all active codes may be balanced to within a range of 20 dB

All figures are without processing and measurement times

The transmit power for the PRACH is set by higher layers based on open loop power control. In the case of DPCH or PUSCH transmission, two or more transport channels may be multiplexed onto a CCTrCH. These transport channels undergo rate matching which involves repetition or puncturing. This rate matching affects the transmit power required to obtain a particular Eb/N0. Thus, the transmission power of the CCTrCH shall be weighted by a gain factor b that either is signalled for the TFC or is computed for the TFC, based upon the signalled settings for a reference TFC. After the synchronization between UTRAN and UE is established, the UE transits into open-loop transmitter power control (TPC).

39

UMTS Air Interface

The power setting for each uplink DPCH in one CCTrCH shall be calculated by the following equation: PUL

aL P

CCPCH

1 ð1

aÞL0 1 IBTS 1 SIRTARGET 1 constant value

ð17Þ

where PUL is the power setting in dBm. This value corresponds to a particular CCTrCH (due to CCTrCH-specific SIRTARGET) and a particular timeslot (due to possibly timeslot-specific a and IBTS). LP-CCPCH is the measure representing path loss in dB (reference transmit power is broadcast on BCH). L0 is the long term average of path loss in dB. IBTS is the interference signal power level at the cell’s receiver in dBm, which is broadcast on BCH. a is a weighting parameter which represents the quality of path loss measurements. a may be a function of the time delay between the uplink time slot and the most recent down link time slot containing a beacon channel. a shall be calculated autonomously at the UE, subject to a maximum allowed value which shall be signalled by higher layers. SIRTARGET is the target SIR in dB. A higher layer outer loop adjusts the target SIR. The constant value shall be set by a higher layer (operator matter) and is broadcast on BCH. If the midamble is used in the evaluation of LP-CCPCH and L0, and the Tx diversity scheme used for the P-CCPCH involves the transmission of different midambles from the diversity antennas, the received power of the different midambles from the different antennas shall be combined prior to evaluation of these variables. The primary CCPCH transmit power is set by higher layer signalling and can be changed based on network determination on a slow basis. The reference transmit power of the PCCPCH is signalled on the BCH. The relative transmit power of the secondary CCPCH and the PICH compared to the PCCPCH transmit power are set by higher layer signalling. The PICH power offset relative to the P-CCPCH reference power is signalled on the BCH. The initial transmission power of the downlink DPCH and the PDSCH is set by the network. After the initial transmission, the UTRAN transits into SIR-based inner loop power control. The UE shall generate TPC commands to control the network transmit power and send them in the TPC field of the uplink DPCH and PUSCH. The association between TPC commands sent on uplink DPCH and PUSCH, with the power controlled downlink DPCH and PDSCH is signalled by higher layers. In the case where no associated downlink data is scheduled within 15 timeslots before the transmission of a TPC command then this is regarded as a transmission pause. The TPC commands in this case shall be derived from measurements on the P-CCPCH. Each TPC command shall always be based on all associated downlink transmissions received since the previous related TPC command. Related TPC commands are defined as TPC commands associated with the same downlink CCTrCHs. If there are no associated downlink transmissions between two or more uplink transmissions carrying related TPC commands, then these TPC commands shall be identical and they shall be regarded by the UTRAN as a single TPC command. This rule applies both to the case where the measurements are based on a CCTrCH or, in the case of a pause, on the P-CCPCH. As a response to the received TPC command, UTRAN may adjust the transmit power. When the TPC command is judged as ‘down’, the transmission power may be reduced by one step, whereas if judged as ‘up’, the transmission power may be raised by one step. The UTRAN may apply an individual offset to the transmission power in each timeslot according to the downlink interference level at the UE. The transmission power of one DPCH or

40

Broadband Wireless Mobile: 3G and Beyond

PDSCH shall not exceed the limits set by higher layer signalling by means of Maximum DL Power (dB) and Minimum DL Power (dB). The transmission power is defined as the average power of the complex QPSK symbols of a single DPCH before spreading. During a downlink transmission pause, the UTRAN may accumulate the TPC commands received. The initial UTRAN transmission power for the first data transmission after the pause may then be set to the sum of transmission power before the pause and a power offset according to the accumulated TPC commands. Additionally this sum may include a constant set by the operator and a correction term due to uncertainties in the reception of the TPC bits. The total downlink transmission power at Node-B within one timeslot shall not exceed maximum transmission power set by higher layer signalling. In the case where the total power of the sum of all transmissions would exceed this limit, then the transmission power of all downlink DPCHs is reduced by the amount that fulfils the requirement. The same amount of power reduction is applied to all DPCHs. A higher layer outer loop adjusts the target SIR. When the dedicated physical channel outof-sync criteria based on the received burst quality is as previously indicated, then the UE shall set the uplink TPC command ‘up’. The CRC-based criteria shall not be taken into account in TPC bit value setting. Timing advanceUTRAN may adjust the UE transmission timing with timing advance. The initial value for timing advance (TAphys) will be determined in the UTRAN by measurement of the timing of the PRACH. The required timing advance will be represented as a 6-bit number (0 63) ‘UL Timing Advance’ TAul, being the multiplier of 4 chips which is nearest to the required timing advance (i.e. TAphys TAul £ 4 chips). When timing advance is used, the UTRAN will continuously measure the timing of a transmission from the UE and send the necessary timing advance value. On receipt of this value the UE shall adjust the timing of its transmissions accordingly in steps of ^4 chips. The transmission of TA values is done by means of higher layer messages. Upon receiving the TA command the UE shall adjust its transmission timing according to the timing advance command at the frame number specified by higher layer signalling. The UE is signalled the TA value in advance of the specified frame activation time to allow for local processing of the command and application of the TA adjustment on the specified frame. Node-B is also signalled the TA value and radio frame number that the TA adjustment is expected to take place. If TA is enabled by higher layers, after handover the UE shall transmit in the new cell with timing advance TA adjusted by the relative timing difference Dt between the new and the old cell: TAnew

TAold 1 2Dt

ð18Þ

If UL synchronization is used, the timing advance is sub-chip granular and with high accuracy in order to enable synchronous CDMA in the UL. The required timing advance will be represented as a multiple of 1/4 chips. The UTRAN will continuously measure the timing of a transmission from the UE and send the necessary timing advance value. On receipt of this value the UE will adjust the timing of its transmissions accordingly in steps of ^1/4 chips. Support of UL synchronization is optional for the UE.

UMTS Air Interface

41

Synchronization procedures During the cell search, the UE searches for a cell and determines the downlink scrambling code, basic midamble code and frame synchronization of that cell. For the dedicated channels, synchronization primitives are used to indicate the synchronization status of radio links, both in uplink and downlink. The definition of the primitives is given next. Layer 1 in the UE shall check the synchronization status in every radio frame of the downlink CCTrC. All bursts and transport channels of a CCTrCH shall be taken into account. Synchronization status is indicated to higher layers, using the CPHY-Sync-IND or CPHYOut-of-Sync-IND primitives. For dedicated physical channels configured with repetition periods only the configured active periods shall be taken into account in the estimation. The status check shall also include detection of the special bursts defined for DTX. Layer 1 in Node B shall check every radio frame synchronization status, individually, for each uplink CCTrCH of the radio link. Synchronization status is indicated to the RL failure/ restored triggering function using either the CPHY-Sync-IND or CPHY-Out-of-Sync-IND primitive. The exact criteria for indicating in-sync/out-of-sync is not subject to specification, but could, for example, be based on received burst quality or CRC checks. One example would be to have the same criteria as for the downlink synchronization status primitives. The downlink CCTrCHs are monitored by the UE, to trigger radio link failure procedures. The downlink CCTrCH failure status is based on the synchronization status primitives CPHY-Sync-IND and CPHY-Out-of-Sync-IND, indicating in-sync and out-of-sync respectively. These primitives shall provide status for each DL CCTrCH separately. The uplink CCTrCHs are monitored by the Node B in order to trigger CCTrCH failure/ restore procedures. The uplink CCTrCH failure/restore status is reported using the synchronization status primitives CPHY-Sync-IND and CPHY-Out-of-Sync-IND, indicating in-sync and out-of-sync, respectively. When the CCTrCH is in the in-sync state, Node B shall start timer T RLFAILURE after receiving N OUTSYNC IND consecutive out-of-sync indications. Node B shall stop and reset timer T RLFAILURE upon receiving successive N INSYNC IND in-sync indications. If T RLFAILURE expires, Node B shall indicate to higher layers which CCTrCHs are out-of-sync using the synchronization status primitives. Furthermore, the CCTrCH state shall be changed to the out-of-sync state. When a CCTrCH is in the out-of-sync state, after receiving N INSYNC IND successive in-sync indications Node B shall indicate that the CCTrCH has re-established synchronization and the CCTrCH’s state shall be changed to the in-sync-state. The specific parameter settings (values of T RLFAILURE, N OUTSYNC IND, and N INSYNC IND) are configurable. Discontinuous transmission (DTX) of radio frames Discontinuous transmission (DTX) is applied in up- and downlink individually for each CCTrCH in case the total bit rate after transport channel multiplexing differs from the total channel bit rate of the dedicated physical channels allocated to a CCTrCH. Rate matching is used in order to fill resource units completely, that are only partially filled with data. In the case where, after rate matching and multiplexing, no data at all is to be transmitted in a resource unit, the complete resource unit is discarded from transmission. This applies also to the case where only one resource unit is allocated and no data has to be transmitted.

42

Broadband Wireless Mobile: 3G and Beyond

In the where case there are no transport blocks provided for transmission by higher layers for any given CCTrCH after link establishment, then a special burst shall be transmitted in the first allocated frame of the transmission pause. If there is a consecutive period of dN OUTSYNC IND/2-1e frames without transport blocks provided by higher layers, then another special burst shall be generated and transmitted at the next possible frame. This pattern shall continue until transport blocks are provided for the CCTrCH by the higher layers. This special burst shall have the same slot format as the burst used for data provided by higher layers. The special burst is filled with an arbitrary bit pattern, contains TFCI and TPC bits if inner loop PC is applied and is transmitted for each CCTrCH individually on the physical channel which is defined to carry the TFCI. The TFCI of the special burst shall indicate that there were no transport blocks provided for transmission by higher layers. The transmission power of the special burst shall be the same as that of the substituted physical channel of the CCTrCH carrying the TFCI. Upon initial establishment and either 160 ms following detection of in-sync, or until the first transport block is received from higher layers, both the UE and the Node B shall transmit the special burst for each CCTrCH for each assigned resource which was scheduled to include a TFCI. Random access procedure The physical random access procedure described below is invoked whenever a higher layer requests transmission of a message on the RACH. The physical random access procedure is controlled by primitives from RRC and MAC. Retransmission on the RACH in case of failed transmission (e.g. due to a collision) is controlled by higher layers. The definition of the RACH in terms of PRACH sub-channels and associated access service classes is broadcast on the BCH in each cell. Parameters for common physical channel uplink outer loop power control are also broadcast on the BCH in each cell. The UE needs to decode this information prior to transmission on the RACH. The physical random access procedure described here is initiated upon request of a PHYData-REQ primitive from the MAC sublayer. Before the physical random-access procedure can be initiated, Layer 1 shall receive the following information by a CPHY-TrCH-Config-REQ from the RRC layer: † the available PRACH sub-channels for each Access Service Class (ASC); † the timeslot, spreading factor, channelization code, midamble, repetition period and offset for each PRACH sub-channel; here is a 1:1 mapping between spreading code and midamble as defined by RRC; † the set of transport format parameters; † the set of parameters for common physical channel uplink outer loop power control.

The physical random-access procedure shall be performed as follows: † Randomly select the PRACH sub-channel from the available ones for the given ASC. The random function shall be such that each of the allowed selections is chosen with equal probability. † Derive the available access slots in the next N frames, defined by SFN, SFN 1 1,…, SFN 1 N 1 for the selected PRACH sub-channel with the help of SFN (where N is the repetition period of the selected PRACH subchannel). Randomly select an uplink access slot from the available access slots in the next frame, defined by SFN, if there is one

UMTS Air Interface

43

available. If there is no access slot available in the next frame, defined by SFN then, randomly select one access slot from the available access slots in the following frame, defined by SFN 1 1. This search is performed for all frames in increasing order, defined by SFN; SFN 1 1; …; SFN 1 N 1, until an available access slot is found. The random function shall be such that each of the allowed selections is chosen with equal probability. † Randomly select a spreading code from the available ones for the given ASC. The random function shall be such that each of the allowed selections is chosen with equal probability. The midamble is derived from the selected spreading code. † Set the PRACH message transmission power level according to the specification for common physical channels in uplink. † Transmit the random access message with no timing advance. DSCH procedure The physical downlink shared channel procedure described below shall be applied by the UE when the physical layer signalling either with the midamble-based signalling or TFCI-based signalling is used to indicate for the UE the need for PDSCH detection. There is also a third alternative to indicate to the UE the need for the PDSCH detection and this is done by means of higher layer signalling. When the UE has been allocated by higher layers to receive data on DSCH using the TFCI, the UE shall decode the PDSCH in the following cases: † In the case of a standalone PDSCH the TFCI is located on the PDSCH itself, then the UE shall decode the TFCI and based on which data rate was indicated by the TFCI, the decoding shall be performed. The UE shall decode PDSCH only if the TFCI word decode corresponds to the TFC part of the TFCS given to the UE by higher layers. † In the case where the TFCI is located on the DCH, the UE shall decode the PDSCH frame or frames if the TFCI on the DCH indicates the need for PDSCH reception. Upon reception of the DCH time slot or time slots, the PDSCH slot (or first PDSCH slot) shall start SFN n 1 2 after the DCH frame containing the TFCI, where n indicates the SFN on which the DCH is received. In the case that the TFCI is repeated over several frames, the PDSCH slot shall start SFN n 1 2 after the frame having the DCH slot which contains the last part of the repeated TFCI.

When the UE has been allocated by higher layers to receive PDSCH based on the midamble used on the PDSCH, the UE shall operate as follows: † The UE shall test the midamble it received and if the midamble received was the same as indicated by higher layers to correspond to PDSCH reception, the UE shall detect the PDSCH data according to the TF given by the higher layers for the UE. † In cases of multiple time slot allocation for the DSCH indicated to be part of the TF for the UE, the UE shall receive all timeslots if the midamble of the first timeslot of PDSCH was the midamble indicated to the UE by higher layers. † If the standalone PDSCH (no associated DCH) contains the TFCI the UE shall detect the TF indicated by the TFCI on PDSCH.

44

Broadband Wireless Mobile: 3G and Beyond

Low chip rate TDD mode The low chip rate TDD mode is under study in the technical report phase, which is well under way. Radio requirements The radio environment recommended by ITU like indoor environment, pedestrian environment, vehicular environment (120 km/h) should be well supported by the low chip rate TDD option. As one option of TDD mode, the low chip rate option should provide the basic service (bearer service). For an IMT-2000 compliant system corresponding to ITU requirement, for the indoor environment, up to 2 Mbps data service should be provided. And for outdoor pedestrian environment, the data service should be up to 384 kbps and more. For the UE in a moving environment (vehicular speed less than 120 km/h), the data rate supported should be 384 or more kbps. Operational requirements The low chip rate TDD option should provide the flexibility to be used for high spot or high density area to provide high speed data service or to provide enhanced coverage or be used alone as a macro cell to provide the service coverage. It should allow deployment together with FDD system, with a high chip rate TDD system, and be similar to a high chip rate TDD deployed with GSM. For the low chip rate TDD option, the deployment should be flexible for all the scenarios such as macro cell, micro cell and pico cell, etc. and also should provide fixed wireless access. For the low chip rate TDD option, the deployment should be flexible for all the scenarios such as macro cell, micro cell and pico cell, etc. and also should provide fixed wireless access. Dependent on the kind of interference accepted by the operator, the operator can vary the maximum cell radius in a trade-off with UL DL interference with the limitations given in Table 2.7. Table 2.7 Interference scenarios and the corresponding max. cell radius Case

Max. cell radius (km)

no UpPTS DwPTS interference allowed UpPTS DwPTS interference allowed, but no interference to TS0 allowed No TS1 DwPTS interference allowed, other interference allowed TS1 DwPTS interference allowed, but no interference to TS0 allowed

11.25 22.5 30 41.25

The guard period of 75 ms between the DwPTS and the UpPTS is designed to avoid interference between the UpPTS (UL) and the DwPTS (DL). Therefore the cell size ensuring the interference-free reception of the DwPTS is guaranteed to a size of approximately 10 km radius (exact value 11.25 km, assuming no delay spread). Consequently, for bigger cell radii there is a conflict in that the advanced UpPTS interferes with the DwPTS reception of another UE close by. Even though the UpPTS DwPTS interference is possible for bigger cell radii then 11.25 km, the impact on the quality of service can be low and acceptable for an operator willing to operate bigger cells.

45

UMTS Air Interface

There are three reasons for that: † The probability that the a UE is close to another UE is low especially for big cells. † The DwPTS needs not to be received by every mobile in every frame. A few DwPTSs not received during initial cell search means no big degradation. The UpPTS is not transmitted every frame; it is only needed for random access or handover. So the probability of disturbance is rather low. It is recommended that the operator avoids interference of TS1 to TS0 by means of the choice of the cell radius. This interference would mean permanent interference for TS0. † The operators can judge the trade-off between quality of service and range and select the range accordingly.

The maximum cell radius dmax is dependent on the time tgap between the potentially interfering UL signal and the potentially interfered DL signal by to following equation: ctgap ð19Þ dmax 2 where c is the velocity of light (Table 2.8). Table 2.8

Allowed cell radius for the occurrence of special UL DL interference

Potentially interfering UL signal

Potentially interfered DL signal

tgap (ms)

dmax (km)

UpPTS UpPTS TS1 TS1

DwPTS TS0 DwPTS TS0

75 150 200 275

11.25 22.5 30 41.25

For the high chip rate option there is no DwPTS guard UpPTS structure. Here, the UL time slots are following the DL time slots immediately. Thus, there is only one step in degree and quality of interference between DL and UL signals. For the low chip rate option, it makes a difference in quality and degree whether the DwPTS or TS is interfered by the UpPTS or TS1. Hence the trade-off between cell range and interference is more manifest for the low chip rate option. The low chip rate TDD option should support the handover between UTRA modes (e.g. low chip rate TDD to high chip rate TDD, low chip rate TDD to FDD), and between systems (e.g. low chip rate TDD to GSM, etc.). Particular characteristics of the low chip rate TDD The features of uplink synchronization, smart antenna (beam forming), etc. have been discussed. These features are to be included in the technical report [42] as they may provide potential performance improvement (Table 2.9). 2.2.1.3 ODMA Although initially proposed as a multi-access scheme, Opportunity Driven Multiple Access(ODMA) must be recognized as a relaying protocol for packetized communication [43].

46

Broadband Wireless Mobile: 3G and Beyond

Table 2.2.9 High level characteristics Parameter/feature

Value/expression

Chip rate Modulation Spreading factor Nominal channel spacing Burst format Radio frame length Sub frame length Time slot number (traffic) Time slot length (ms) Downlink pilot slot (ms) Uplink pilot slot (ms) Guard period (ms) Range of uplink slot Range of downlink slot Receiver type Pilot aided detection Synchronization aspect Precision for UL sync. Antenna processing Switching point Power control/rate Variable bit rate service Basic resource unit

1.28 Mcps QPSK (8PSK) 1/2/4/8/16 1.6 MHz /carrier 1 burst type 10 ms (divided into 2 sub frames) 5 ms 7 675 75 125 75 1 6 1 6 Multi user detection (option), RAKE DwPTS, UpPTS, Midamble Downlink and uplink synchronization 1/8 chip Smart antenna with beam forming Two switching points/sub frame Open loop power control; closed loop power control/200 Hz (max rate) Supported (using TFCI) One code, one slot with spreading factor ¼ 16 (use of same resource in both consecutive sub frames) Multi code, multi slot combination (variable spreading factor) 10/20/40/80 ms Low chip rate TDD to High chip rate TDD, FDD, GSM, etc. Same capability as high chip rate TDD for DwPTS, DPCH, but not for P CCPCH; the Tx Diversity scheme used for the DPCH is used for the FPACH, as well

Service mapping Interleaving period HO capability Tx diversity

Basically it is a mode of operation of UE that permits the relay of communication through other UEs to reach the target BS by breaking the direct path to Node B into smaller paths. This mode of operation is not feasible in FDD mode as it would require a UE capable of transmitting and receiving in the same frequency band. Thus ODMA is only viewed as possible within the TDD mode. How it works When the UE is switched on, it builds a database of the nearest five neighbours by probing. It sends packets with a predetermined power and awaits response for a period of time. Once

UMTS Air Interface

47

the response is received (ACK) the UE is added in the database of neighbours as being available. When a response is not received, the power of the probing packet is increased and the time of response is reset. After the database is built, the UE periodically probes the accuracy of this database (dumping disappearing neighbours and adding new ones) by probing. This probe also provides information on its nearest neighbours. Thus each UE has double-hop information, meaning that it has the necessary information to instruct a packet to make a definite double hop. When a UE requires to send any packet, the information stored in its database will be enough to select the best path to Node B from a pool of paths. Benefits of ODMA The proposal of ODMA has the main advantage of any radio relaying system. Breaking a direct path into smaller ones overcomes the path loss more and more efficiently as the number of relaying stations is increased. In addition to this advantage, the use of ODMA has other potential gains related to the cellular nature of UMTS: † Relaying to avoid shadowing: by use of relaying stations, shadowing can be avoided by using the update period of the nearest neighbour’s database. The method is straightforward. The appearance of fading can be overcome by using an alternative path of relaying UEs, thus increasing the signal-to-noise ratio dramatically. For example, in a ‘Manhattan’ environment where sharp turns and high buildings make the deployment of a high number of base stations mandatory, this mode of operation is clearly an option. † Radio resource reuse: due to the lower power of the UEs, the reuse of resources (such as timeslots and codes) is theoretically possible, instead of invalidating a certain timeslot or a code in a whole cell, due to the smaller ‘footprint’ of the UE. † Potential for capacity gain: also due to the smaller ‘footprint’ of UEs, it is possible to reduce the inter-cell interference. In order to be able to cover the maximum surface, cells are bound to overlap. This is the main cause of inter-cell interference that is responsible for a great deal of the degradation of the signal in UMTS. Using a smaller cell ‘footprint’ and due to the relaying capability of the UEs, users are not to suffer such high inter-cell interference.

The capability for managing the so-called ‘hot spots’ is evident. If a user finds that the path leading to the closest Node B is saturated, it is possible to find a still valid Node B by relaying to another path. Another valid technique to deal with the hot spots is by deploying high-capacity UEs (e.g. 24 dBm terminals) to a certain cell in order to make the transition smoother to other not-so-saturated cells. This technique has great potential as it is not so costly and is much faster than deploying a new Node B to deal with a pico cell environment. † Resilience: one of the greatest advantages of the Internet is that, due to the capability of relaying and finding different paths, it enjoys a high degree of resilience, i.e. it is resistant to the loss of connectivity between (any) two given nodes. As the breaking of the path into smaller paths gain the bonus of lower path loss, the increment in the number of relaying UEs would give UMTS an additional resistance to loss of communication.

48

Broadband Wireless Mobile: 3G and Beyond

2.2.2 Layer 2 2.2.2.1 MAC MAC architecture According to the RRC functions (see Section 2.3) the RRC is generally in control of the internal configuration of the MAC [36]. The diagrams that describe the MAC architecture are constructed from MAC entities. The entities are assigned the following names: † MAC-b is the MAC entity that handles the following transport channels:

broadcast channel (BCH). † MAC-c/sh, is the MAC entity that handles the following transport channels:

paging channel (PCH); forward access channel (FACH); random access channel (RACH); common packet channel (UL CPCH); the CPCH exists only in FDD mode; downlink shared channel (DSCH); uplink shared channel (USCH); the USCH exists only in TDD mode. † MAC-d is the MAC entity that handles the following transport channels:

dedicated transport channels (DCH). The exact functions completed by the entities are different in the UE from those completed in the UTRAN. Figure 2.13 illustrates the connectivity of the MAC-b entity in a UE and in each cell of the UTRAN. MAC-b represents the control entity for the broadcast channel (BCH). There is one MAC-b entity in each UE and one MAC-b in the UTRAN for each cell. The MAC Control SAP is used to transfer Control information to MAC-b. The MAC-b entity is located in the Node B. Figure 2.14 illustrates the connectivity of MAC entities. The MAC-c/sh controls access to common transport channels. The MAC-d controls access to dedicated transport channels. If logical channels of dedicated type are mapped to common channels then MAC-d passes the data to MAC-c/sh via the illustrated connection between the functional entities. The mapping of logical channels on transport channels depends on the multiplexing that is configured by RRC. The MAC Control SAP is used to transfer control information to each MAC entity. Figure 2.14 shows the UE side MAC-c/sh entity. The following functionality is covered: † TCTF MUX: this function represents the handling (insertion for uplink channels and detection and deletion for downlink channels) of the TCTF field in the MAC header, and the respective mapping between logical and transport channels; the TCTF field indicates the common logical channel type, or if a dedicated logical channel is used. † Add/read UE Id: the UE Id is added for CPCH and RACH transmissions; the UE Id, when present, identifies data to this UE.

UMTS Air Interface

49

Figure 2.13 UE side and UTRAN side architecture. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

Figure 2.14 UE side MAC architecture. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

50

Broadband Wireless Mobile: 3G and Beyond

† UL: TF selection: in the uplink, the possibility of transport format selection exists; in the case of CPCH transmission, a TF is selected based on TF availability determined from status information on the CSICH. † ASC selection: for RACH, MAC indicates the ASC associated with the PDU to the physical layer. For CPCH, MAC may indicate the ASC associated with the PDU to the physical layer. This is to ensure that RACH and CPCH messages associated with a given Access Service Class (ASC) are sent on the appropriate signature(s) and time slot(s). MAC also applies the appropriate back-off parameter(s) associated with the given ASC. † Scheduling /priority handling: this functionality is used to transmit the information received from MAC-d on RACH and CPCH based on logical channel priorities. This function is related to TF selection. † TFC selection: transport format and transport format combination selection according to the transport format combination set (or transport format combination subset) configured by RRC is performed.

The RLC provides RLC-PDUs to the MAC, which fit into the available transport blocks on the transport channels. There is one MAC-c/sh entity in each UE. The following functionality is covered by the MAC-d entity at the UE side: † Channel switching: dynamic transport channel type switching is performed by this entity, based on decision taken by RRC. This is usually related to a change of radio resources. † C/T MUX: the C/T MUX is used when multiplexing of several dedicated logical channels onto one transport channel is used. An unambiguous identification of the logical channel is included. † Ciphering: Ciphering for transparent mode data to be ciphered is performed in MAC-d. † Deciphering: Deciphering for ciphered transparent mode data is performed in MAC-d. † UL TFC selection: transport format and transport format combination selection according to the transport format combination set (or transport format combination subset) configured by RRC is performed.

The MAC-d entity is responsible for mapping dedicated logical channels for the uplink either onto dedicated transport channels or to transfer data to MAC-c/sh to be transmitted via common channels. One dedicated logical channel can be mapped simultaneously onto DCH and DSCH. The MAC-d entity has a connection to the MAC-c/sh entity. This connection is used to transfer data to the MAC-c/sh to transmit data on transport channels that are handled by MAC-c/sh (uplink) or to receive data from transport channels that are handled by MAC-c/ sh (downlink). There is one MAC-d entity in the UE. Figure 2.15 illustrates the connectivity between the MAC entities from the UTRAN side. It is similar to the UE case with the exception that there will be one MAC-d for each UE and each UE (MAC-d) that is associated with a particular cell may be associated with that cell’s MAC-c/sh. MAC-c/sh is located in the controlling RNC while MAC-d is located in the serving RNC. The MAC Control SAP is used to transfer control information to each MAC entity belongs to one UE. The following functionality is covered by the MAC-c/sh entity at the UTRAN side: † The scheduling priority handling: this function manages FACH and DSCH resources between the UE’s and between data flows according to their priority. † TCTF MUX: this function represents the handling (insertion for downlink channels and detection and deletion for uplink channels) of the TCTF field in the MAC header, and the

UMTS Air Interface

51

Figure 2.15 UTRAN side MAC architecture. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

† † † †

respective mapping between logical and transport channels. The TCTF field indicates the common logical channel type, or if a dedicated logical channel is used. UE Id Mux: for dedicated type logical channels, the UE Id field in the MAC header is used to distinguish between UEs. TFC selection: in the downlink, transport format combination selection is done for FACH and PCH and DSCHs. Demultiplex: for TDD operation the demultiplex function is used to separate USCH data from different UEs, i.e. to be transferred to different MAC-d entities. DL code allocation: this function is used to indicate the code used on the DSCH.

Flow control is provided to MAC-d. The RLC provides RLC-PDUs to the MAC, which fit into the available transport blocks on the transport channels. There is one MAC-c/sh entity in the UTRAN for each cell. The following functionality is covered by the MAC-d entity at the UTRAN side: † Channel switching: dynamic transport channel type switching is performed by this entity, based on decision taken by RRC. † C/T MUX box: the function includes the C/T field when multiplexing of several dedicated logical channels onto one transport channel is used. † Priority setting: this function is responsible for priority setting on data received from DCCH / DTCH. † Ciphering: ciphering for transparent mode data to be ciphered is performed in MAC-d. † Deciphering: deciphering for ciphered transparent mode data is performed in MAC-d. † DL scheduling/priority handling: in the downlink, scheduling and priority handling of

52

Broadband Wireless Mobile: 3G and Beyond

transport channels is performed within the allowed transport format combinations of the TFCS assigned by the RRC. † Flow control: a flow control function exists toward MAC-c/sh to limit buffering between MAC-d and MAC-c/sh entities. This function is intended to limit Layer 2 signalling latency and reduce discarded and retransmitted data as a result of FACH or DSCH congestion. A MAC-d entity using common channels is connected to a MAC-c/sh entity that handles the scheduling of the co mmon channels to which the UE is assigned and DL (FACH) priority identification to MAC-c/sh. A MAC-d entity using downlink shared channel is connected to a MAC-c/sh entity that handles the shared channels to which the UE is assigned and indicates the level of priority of each PDU to MAC-c/sh. A MAC-d entity is responsible for mapping dedicated logical channels onto the available dedicated transport channels or routing the data received on a DCCH or DTCH to MAC-c/sh. One dedicated logical channel can be mapped simultaneously on DCH and DSCH. Different scheduling mechanisms apply for DCH and DSCH. There is one MAC-d entity in the UTRAN for each served UE. MAC services and functionsServices and functions provided by the MAC sublayer to upper layers are: † Data transfer: this service provides unacknowledged transfer of MAC PDUs between peer MAC entities. This service does not provide any data segmentation. Therefore, segmentation/reassemble function should be achieved by upper layer. † Reallocation of radio resources and MAC parameters: this service performs on request of RRC execution of radio resource reallocation and change of MAC parameters, i.e. reconfiguration of MAC functions such as change of identity of UE, change of transport format (combination) sets, change of transport channel type. In TDD mode, in addition, the MAC can handle resource allocation autonomously. † Reporting of measurements: local measurements such as traffic volume and quality indication are reported to RRC.

Logical channels The MAC layer provides data transfer services on logical channels (Figure 2.10). A set of logical channel types is defined for different kinds of data transfer services as offered by MAC. Each logical channel type is defined by what type of information is transferred. As already described, a general classification of logical channels is in two groups: † control channels (for the transfer of control plane information). † traffic channels (for the transfer of user plane information).

Control channels are used for transfer of control plane information only. † Broadcast Control Channel (BCCH): a downlink channel for broadcasting system control information. † Paging Control Channel (PCCH): a downlink channel that transfers paging information. This channel is used when the network does not know the location cell of the UE, or, the UE is in the cell connected state (utilizing UE sleep mode procedures). † Common Control Channel (CCCH): bi-directional channel for transmitting control

UMTS Air Interface

† † † †

53

information between network and UEs. This channel is commonly used by the UEs having no RRC connection with the network and by the UEs using common transport channels when accessing a new cell after cell reselection. Dedicated Control Channel (DCCH): a point-to-point bi-directional channel that transmits dedicated control information between a UE and the network. This channel is established through RRC connection setup procedure. Shared Channel Control Channel (SHCCH): bi-directional channel that transmits control information for uplink and downlink shared channels between network and UEs. This channel is for TDD only. ODMA Common Control Channel (OCCCH): bi-directional channel for transmitting control information between UEs. ODMA Dedicated Control Channel (ODCCH): a point-to-point bi-directional channel that transmits dedicated control information between UEs. This channel is established through RRC connection setup procedure. Traffic channels are used for the transfer of user plane information only.

† Dedicated Traffic Channel (DTCH): a point-to-point channel, dedicated to one UE, for the transfer of user information. A DTCH can exist in both uplink and downlink. † ODMA Dedicated Traffic Channel (ODTCH): a point-to-point channel, dedicated to one UE, for the transfer of user information between UEs. An ODTCH exists in relay link. † Common Traffic Channel (CTCH): a point-to-multipoint unidirectional channel for transfer of dedicated user information for all or a group of specified UEs.

The following connections between logical channels and transport channels exist: † † † † †

BCCH is connected to BCH and may also be connected to FACH; PCCH is connected to PCH; CCCH is connected to RACH and FACH; SHCCH is connected to RACH and USCH/FACH and DSCH; DTCH can be connected to either RACH and FACH, to RACH and DSCH, to DCH and DSCH, to a DCH, a CPCH (FDD only) or to USCH (TDD only); † CTCH is connected to FACH; † DCCH can be connected to either RACH and FACH, to RACH and DSCH, to DCH and DSCH, to a DCH, a CPCH (FDD only) to FAUSCH, CPCH (FDD only), or to USCH (TDD only). The mappings shown in Figures 2.16 and 2.17 illustrates the mapping from the UE in relay operation. Note that ODMA logical channels and transport channels are employed only in relay link transmissions (i.e. not used for uplink or downlink transmissions on the UEUTRAN radio interface). MAC functions

The functions of MAC include:

† Mapping between logical channels and transport channels. The MAC is responsible for mapping of logical channel(s) onto the appropriate transport channel(s). † Selection of appropriate transport format for each transport channel depending on instantaneous source rate. Given the transport format combination set assigned by RRC, MAC selects the appropriate transport format within an assigned transport format set for each

54

Broadband Wireless Mobile: 3G and Beyond

Figure 2.16

Logical channels mapped onto transport channel.

active transport channel depending on source rate. The control of transport formats ensures efficient use of transport channels. † Priority handling between data flows of one UE. When selecting between the transport format combinations in the given transport format combination set, priorities of the data flows to be mapped onto the corresponding transport channels can be taken into account. Priorities are, e.g. given by attributes of radio bearer services and RLC buffer status. The priority handling is achieved by selecting a transport format combination for which high priority data is mapped onto L1 with a ‘high bit rate’ transport format, at the same time letting lower priority data be mapped with a ‘low bit rate’ (could be zero bit rate) transport format. Transport format selection may also take into account transmit power indication from Layer 1. † Priority handling between UEs by means of dynamic scheduling. In order to utilize the spectrum resources efficiently for bursty transfer, a dynamic scheduling function may be

Figure 2.17

Logical channels mapped onto transport channel (ODMA mode only).

UMTS Air Interface

†

†

†

† † † †

55

applied. MAC realizes priority handling on common and shared transport channels. Note that for dedicated transport channels, the equivalent of the dynamic scheduling function is implicitly included as part of the reconfiguration function of the RRC sublayer. Identification of UEs on common transport channels. When a particular UE is addressed on a common downlink channel, or when a UE is using the RACH, there is a need for inband identification of the UE. Since the MAC layer handles the access to, and multiplexing onto, the transport channels, the identification functionality is naturally also placed in MAC. Multiplexing/demultiplexing of higher layer PDUs into/from transport blocks delivered to/from the physical layer on common transport channels. MAC should support service multiplexing for common transport channels, since the physical layer does not support multiplexing of these channels. Multiplexing/demultiplexing of higher layer PDUs into/from transport block sets delivered to/from the physical layer on dedicated transport channels. The MAC allows service multiplexing for dedicated transport channels. This function can be utilized when several upper layer services (e.g. RLC instances) can be mapped efficiently on the same transport channel. In this case the identification of multiplexing is contained in the MAC protocol control information. Traffic volume monitoring. Measurement of traffic volume on logical channels and reporting to RRC. Based on the reported traffic volume information, RRC performs transport channel switching decisions. Dynamic Transport Channel type switching. Execution of the switching between common and dedicated transport channels based on a switching decision derived by RRC. Ciphering. This function prevents unauthorized acquisition of data. Ciphering is performed in the MAC layer for transparent RLC mode. Access service class selection for RACH transmission. The RACH resources (i.e. access slots and preamble signatures for FDD, timeslot and channelization code for TDD) may be divided between different access service classes in order to provide different priorities of RACH usage. In addition it is possible for more than one ASC or for all ASCs to be assigned to the same access slot/signature space. Each access service class will also have a set of back-off parameters associated with it, some or all of which may be broadcast by the network. The MAC function applies the appropriate back-off and indicates to the PHY layer the RACH partition associated to a given MAC PDU transfer.

RLC RLC architecture Figure 2.18 gives an overview model of the RLC layer. The figure illustrates the different RLC peer entities. There is one transmitting and one receiving entity for the transparent mode service and the unacknowledged mode service and one combined transmitting and receiving entity for the acknowledged mode service. In [37] the word transmitted is equivalent to ‘submitted to lower layer’ unless otherwise explicitly stated. The dashed lines between the AM-Entities illustrate the possibility to send the RLC PDUs on separate logical channels, e.g. control PDUs on one and data PDUs on the other. Transparent mode entities

The transmitting Tr-entity receives SDUs from the higher layers

56

Broadband Wireless Mobile: 3G and Beyond

Figure 2.18 Overview model of RLC. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

UMTS Air Interface

57

through the Tr-SAP. RLC might segment the SDUs into appropriate RLC PDUs without adding any overhead. How to perform the segmentation is decided upon when the service is established. RLC delivers the RLC PDUs to MAC through either a BCCH, DCCH, PCCH, SHCCH or a DTCH. The CCCH and SHCCH also uses transparent mode, but only for the uplink. Which type of logical channel is used depends on whether the higher layer is located in the control plane (BCCH, DCCH, PCCH, CCCH, SHCCH) or user plane (DTCH). The Tr-entity receives PDUs through one of the logical channels from the MAC sublayer. RLC reassembles (if segmentation has been performed) the PDUs into RLC SDUs. How to perform the reassembling is decided upon when the service is established. RLC delivers the RLC SDUs to the higher layer through the Tr-SAP. Unacknowledged mode entities The transmitting UM-entity receives SDUs from the higher layers. RLC might segment the SDUs into RLC PDUs of appropriate size. The SDU might also be concatenated with other SDUs. RLC delivers the RLC PDUs to MAC through either a DCCH, CTCH or a DTCH. The CCCH and SHCCH also uses unacknowledged mode, but only for the downlink. Which type of logical channel used depends on whether the higher layer is located in the control plane (CCCH, DCCH, SHCCH) or user plane (CTCH, DTCH). The receiving UM-entity receives PDUs through one of the logical channels from the MAC sublayer. RLC removes headers from the PDUs and reassembles the PDUs (if segmentation has been performed) into RLC SDUs. The RLC SDUs are delivered to the higher layer. Acknowledged mode entity In the situation where two logical channels are used in the uplink the first logical channel shall be used for data PDUs and the second logical channel shall be used for control PDUs. Where one logical channel is used, the RLC PDU size shall be the same for AMD PDUs and control PDUs. The transmitting side of the AM-entity receives SDUs from the higher layers. The SDUs are segmented and/or concatenated to PUs of fixed length. PU length is a semi-static value that is decided in bearer setup and can only be changed through bearer reconfiguration by RRC. For purposes of RLC buffering and retransmission handling, the operation is the same as if there would be one PU per PDU. For concatenation or padding purposes, bits of information on the length and extension are inserted into the beginning of the last PU where data from an SDU is included. Padding can be replaced by piggybacked status information. This includes setting the poll bit. If several SDUs fit into one PU, they are concatenated and the appropriate length indicators are inserted into the beginning of the PU. After that the PUs are placed in the retransmission buffer and the transmission buffer. One PU is included in one RLC PDU. The MUX then decides which PDUs and when the PDUs are submitted to a lower layer. The PDUs are submitted via a function that completes the RLC-PDU header and potentially replaces padding with piggybacked status information. The RLC entity shall assume a PDU to be transmitted when the PDU is submitted to lower layer. The ciphering is applied only for AMD PDUs. The fixed 2-octet AMD PDU header is not ciphered. Piggybacked and padding parts of AMD PDU, when existing, are ciphered. The other Control PDUs (e.g. STATUS, RESET, and RESET ACK PDU) shall not be ciphered. When the piggybacking mechanism is applied, the padding is replaced by control information, in order to increase the transmission efficiency making possible a faster message exchange between the peer to peer RLC entities. The piggybacked control information is

58

Broadband Wireless Mobile: 3G and Beyond

not saved in any retransmission buffer. The piggybacked control information is contained in the piggybacked STATUS PDU, which is in turn included into the AMD-PDU. The piggybacked STATUS PDUs will be of variable size in order to match with the amount of free space in the AMD PDU. The retransmission buffer also receives acknowledgements from the receiving side, which are used to indicate retransmissions of PUs and when to delete a PU from the retransmission buffer. The receiving side of the AM-entity receives PDUs through one of the logical channels from the MAC sublayer. The RLC-PDUs are expanded into separate PUs and potential piggybacked status information is extracted. The PUs are placed in the receiver buffer until a complete SDU has been received. The receiver buffer requests retransmissions of PUs by sending negative acknowledgements to the peer entity. After that the headers are removed from the PDUs and the PDUs are reassembled into a SDU. Finally the SDU is delivered to the higher layer. The receiving side also receives acknowledgements from the peer entity. The acknowledgements are passed to the retransmission buffer on the transmitting side. RLC services and functions The services provided to the upper layer are as follows: † RLC connection establishment/release: this service performs establishment/release of RLC connections. † Transparent data transfer: this service transmits higher layer PDUs without adding any protocol information, possibly including segmentation/reassemble functionality. † Unacknowledged data transfer: this service transmits higher layer PDUs without guaranteeing delivery to the peer entity. The unacknowledged data transfer mode has the following characteristics: † SUBDetection of erroneous data: the RLC sublayer shall deliver only those SDUs to the receiving higher layer that are free of transmission errors by using the sequence-number check function; † SUBThe RLC sublayer shall deliver each SDU only once to the receiving upper layer using duplication detection function; † SUBThe receiving RLC sublayer entity shall deliver a SDU to the higher layer receiving entity as soon as it arrives at the receiver. † Acknowledged data transfer. This service transmits higher layer PDUs and guarantees delivery to the peer entity. In case RLC is unable to deliver the data correctly, the user of RLC at the transmitting side is notified. For this service, both in-sequence and out-ofsequence delivery are supported. In many cases a higher layer protocol can restore the order of its PDUs. As long as the out-of-sequence properties of the lower layer are known and controlled (i.e. the higher layer protocol will not immediately request retransmission of a missing PDU) allowing out-of-sequence delivery can save memory space in the receiving RLC. The acknowledged data transfer mode has the following characteristics: † SUBError-free delivery: error-free delivery is ensured by means of retransmission. The receiving RLC entity delivers only error-free SDUs to the higher layer. † SUBUnique delivery: the RLC sublayer shall deliver each SDU only once to the receiving upper layer using duplication detection function. † SUBIn-sequence delivery: RLC sublayer shall provide support for in-order delivery of

UMTS Air Interface

59

SDUs, i.e. RLC sublayer should deliver SDUs to the receiving higher layer entity in the same order as the transmitting higher layer entity submits them to the RLC sublayer. † SUBOut-of-sequence delivery: as an alternative to in-sequence delivery, it shall also be possible to allow that the receiving RLC entity delivers SDUs to higher layer in different order than submitted to RLC sublayer at the transmitting side. † QoS setting: the retransmission protocol shall be configurable by Layer 3 to provide different levels of QoS. This can be controlled. † Notification of unrecoverable errors:RLC notifies the upper layer of errors that cannot be resolved by RLC itself by normal exception handling procedures, e.g. by adjusting the maximum number of retransmissions according to delay requirements. There is a single RLC connection per radio bearer. The RLC functionality is described in the following items: † Segmentation and reassemble: this function performs segmentation/reassemble of variable-length higher layer PDUs into/from smaller RLC Payload Units (PUs). The RLC PDU size is adjustable to the actual set of transport formats. † Concatenation: if the contents of an RLC SDU do not fill an integer number of RLC PUs, the first segment of the next RLC SDU may be put into the RLC PU in concatenation with the last segment of the previous RLC SDU. † Padding: when concatenation is not applicable and the remaining data to be transmitted does not fill an entire RLC PDU of given size, the remainder of the data field shall be filled with padding bits. † Transfer of user data: this function is used for conveyance of data between users of RLC services. RLC supports acknowledged, unacknowledged and transparent data transfer. QoS setting controls transfer of user data. † Error correction: this function provides error correction by retransmission (e.g. Selective Repeat, Go Back N, or a Stop-and-Wait ARQ) in acknowledged data transfer mode. † In-sequence delivery of higher layer PDUs: this function preserves the order of higher layer PDUs that were submitted for transfer by RLC using the acknowledged data transfer service. If this function is not used, out-of-sequence delivery is provided. † Duplicate detection: this function detects duplicated received RLC PDUs and ensures that the resultant higher Layer PDU is delivered only once to the upper layer. † Flow control: this function allows an RLC receiver to control the rate at which the peer RLC transmitting entity may send information. † Sequence number check (unacknowledged data transfer mode): this function guarantees the integrity of reassembled PDUs and provides a mechanism for the detection of corrupted RLC SDUs through checking sequence number in RLC PDUs when they are reassembled into a RLC SDU. A corrupted RLC SDU will be discarded. † Protocol error detection and recovery: this function detects and recovers from errors in the operation of the RLC protocol. † Ciphering: this function prevents unauthorized acquisition of data. Ciphering is performed in RLC layer for non-transparent RLC mode. † Suspend/resume function: suspension and resumption of data transfer as in e.g. LAPDm (cf. GSM 04.05).

60

Broadband Wireless Mobile: 3G and Beyond

PDCP PDCP architecture Figure 2.19 shows the model of the Packet Data Convergence Protocol (PDCP) [38] within the UTRAN protocol architecture. Every PDCP-SAP uses exactly one PDCP entity. Each PDCP entity uses none, one or several header compression algorithm types.

Figure 2.19 PDCP structure. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

PDCP services and functions † PDCP services provided to upper layers: † SUBTransmission and reception of network PDUs in acknowledged, unacknowledged and transparent RLC mode. † PDCP functions: † SUBMapping of Network PDUs from one network protocol to one RLC entity. † SUBCompression in the transmitting entity and decompression in the receiving entity of redundant Network PDU control information (header compression/ decompression). This may include TCP/IP header compression and decompression.

UMTS Air Interface

61

BMC BMC architecture Broadcast/Multicast Control (BMC) [39] is a sublayer of L2 that exists in the user-plane only. It is located above RLC. The L2/BMC sublayer is assumed to be transparent for all services except broadcast/multicast. Figure 2.20 shows the model of the L2/BMC sublayer within the UTRAN radio interface protocol architecture.

Figure 2.20 BMC protocol model. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

At the UTRAN side, the BMC sublayer shall consist of one BMC protocol entity per cell. Each BMC entity requires a single CTCH, which is provided by the MAC sublayer, through the RLC sublayer. The BMC requests the unacknowledged mode service of the RLC. It is assumed that there is a function in the RNC above BMC that resolves the geographical area information of the CB message (or, if applicable, performs evaluation of a cell list) received from the Cell Broadcast Centre (CBC). A BMC protocol entity serves only those messages at BMC-SAP that are to be broadcast into a specified cell. Broadcast/multicast control services and functions This section provides an overview on services and functions provided by the BMC sublayer.

62

Broadband Wireless Mobile: 3G and Beyond

BMC services The BMC provides a broadcast/multicast transmission service in the user plane on the radio interface for common user data in transparent or unacknowledged mode. BMC functions † Storage of cell broadcast messages: the BMC stores the cell broadcast messages received over the CBC-RNC interface for scheduled transmission. † Traffic volume monitoring and radio resource request for CBS: at the UTRAN side, the BMC calculates the required transmission rate for cell broadcast service based on the messages received over the CBC RNC interface, and requests for appropriate CTCH/ FACH resources from RRC. † Scheduling of BMC messages: the BMC receives scheduling information together with each cell broadcast message over the CBC RNC-interface. Based on this scheduling information, at the UTRAN side, BMC generates schedule messages and schedules BMC message sequences accordingly. At the UE side, BMC evaluates the schedule messages and indicates scheduling parameters to RRC, which are used by RRC to configure the lower layers for CBS discontinuous reception. † Transmission of BMC messages to UE: this function transmits the BMC messages (scheduling and cell broadcast messages) according to schedule. † Delivery of Cell Broadcast messages to upper layer (NAS): this function delivers the received cell broadcast messages to upper layer (NAS) in the UE. Only non-corrupted cell broadcast messages are delivered.

2.2.3 Layer 3 The network layer in UTRAN is divided into Control (C-) and User (U-) planes [10]. User plane protocols transport user data through the access stratum while control plane protocols control the radio access bearers and the connection between the UE and the network. In the C-plane, Layer 3 is split into the following sub-layers: Radio Resource Control (RRC) [40] and duplication avoidance. RRC interfaces with Layer 2 and duplication avoidance provides the access stratum services to higher layers. The higher Layer 3 functions such as mobility management and call control are assumed to belong to the non-access stratum and therefore they are not described in this section. The interface between duplication avoidance and higher Layer 3 sub-layers is defined by the General Control (GC), Notification (Nt) and Dedicated Control (DC) Service Access Points (SAP). Figure 2.21 shows the situation of Layer 3 and its two sub-layers in the context of UMTS Radio Interface Protocol Architecture. Service Access Points (SAP) are marked with circles at the interface between sub-layers. Also shown in this figure are connections between RRC and lower layers (PHY, MAC) providing inter-layer control services. Access stratum services provided by Layer 3 Layer 3 provides the following services in the access stratum of UMTS radio interface [41] through the corresponding service access points: general control, notification and dedicated control.

UMTS Air Interface

63

Figure 2.21 Situation of Layer 3 in UMTS Radio Interface Protocol Architecture. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

General control General Control (GC) provides an information broadcast service to all UE in a certain geographical area. The basic requirements for this service are: † It should be possible to broadcast non-access stratum information in a certain geographical area. † The information is transferred on an unacknowledged mode link, i.e. the delivery of the broadcast information cannot be guaranteed (typically no retransmission scheme is used). † It should be possible to do repeated transmissions of the broadcast information (controlled by the non-access stratum).

The point where the UE receives the broadcast information should be included when the access stratum delivers broadcast information to the non-access stratum.

64

Broadband Wireless Mobile: 3G and Beyond

Notification Notification (Nt) provides paging and notification broadcast services. The information is broadcast in a certain geographical area but it is only addressed to one or several specific UEs. The basic requirements for the paging service are: † It should be possible to broadcast paging information to a number of UEs in a certain geographical area. † The information is transferred on an unacknowledged mode link. It is assumed that the protocol entities in non-access stratum handle any kind of retransmission of paging information. † The requirements for the notification broadcast services are similar to those of the information broadcast service of the GC. † It should be possible to broadcast notification information in a certain geographical area. † The information is transferred on an unacknowledged mode link.

Dedicated control Dedicated Control (DC) provides services for establishment, transfer of messages and release of a connection. It should be possible to transfer a message not only using an existing connection but also during the establishment phase. The basic requirements for the establishment/release services are: † It should be possible to establish connections (both point and group connections). † It should be possible to transfer an initial message during the connection establishment phase. This message transfer has the same requirements as the information transfer service. † It should be possible to release connections.

The information transfer service allows the specification of the quality of service requirements for each message. The basic requirements from the information transfer service provided by the duplication avoidance function are: † In-sequence transfer of messages: messages are delivered to the Non-Access Stratum (NAS) on the receiver side exactly in the order they have been submitted by the NAS on the sending side, without loss or duplication, except possibly for the loss of last messages in case of connection abortion. † Priority handling: if SMS messages should be transported through the control plane it should be possible to give higher priority to signalling messages.

Radio resource control functions and entities The Radio Resource Control (RRC) sub-layer [40] handles the control plane signalling of Layer 3 between the UE and UTRAN. It performs the following functions: † Broadcast of information provided by the non-access stratum (core network): the RRC layer performs system information broadcasting from the network to all UEs. The system information is normally repeated on a regular basis. The RRC layer performs the scheduling, segmentation and repetition. This function supports broadcast of higher layer (above RRC) information. This information may be cell-specific or not. As an example

UMTS Air Interface

†

†

†

†

†

† † †

†

65

RRC may broadcast core network location service area information related to some specific cells. Broadcast of information related to the access stratum: the RRC layer performs system information broadcasting from the network to all UEs. The system information is normally repeated on a regular basis. The RRC layer performs the scheduling, segmentation and repetition. This function supports broadcast of typically cell-specific information. Establishment, maintenance and release of an RRC connection between the UE and UTRAN: the establishment of an RRC connection is initiated by a request from higher layers at the UE side to establish the first signalling connection for the UE. The establishment of an RRC connection includes an optional cell re-selection, an admission control, and a Layer 2 signalling link establishment. The release of an RRC connection can be initiated by a request from higher layers to release the last signalling connection for the UE or by the RRC layer itself in case of RRC connection failure. In the case of connection loss, the UE requests re-establishment of the RRC connection. In the case of RRC connection failure, RRC releases resources associated with the RRC connection. Establishment, reconfiguration and release of radio bearers: the RRC layer can, on request from higher layers, perform the establishment, reconfiguration and release of radio bearers in the user plane. A number of radio bearers can be established to a UE at the same time. At establishment and reconfiguration, the RRC layer performs admission control and selects parameters describing the radio bearer processing in Layer 2 and Layer 1, based on information from higher layers. Assignment, reconfiguration and release of radio resources for the RRC connection: the RRC layer handles the assignment of radio resources (e.g. codes, CPCH channels) needed for the RRC connection including needs from both the control and user plane. The RRC layer may reconfigure radio resources during an established RRC connection. This function includes coordination of the radio resource allocation between multiple radio bearers related to the same RRC connection. RRC controls the radio resources in the uplink and downlink such that UE and UTRAN can communicate using unbalanced radio resources (asymmetric uplink and downlink). RRC signals to the UE to indicate resource allocations for purposes of handover to GSM or other radio systems. RRC connection mobility functions: the RRC layer performs evaluation, decision and execution-related to RRC connection mobility during an established RRC connection, such as handover, preparation of handover to GSM or other systems, cell re-selection and cell/paging area update procedures, based on, for example, measurements done by the UE. Routing of higher layer PDUs: this function performs at the UE side routing of higher layer PDUs to the correct higher layer entity, and at the UTRAN side to the correct RANAP entity. Control of requested QoS: this function shall ensure that the QoS requested for the radio bearers can be met. This includes the allocation of a sufficient number of radio resources. UE measurement reporting and control of the reporting: measurements performed by the UE are controlled by the RRC layer, in terms of what to measure, when to measure and how to report, including both UMTS air interface and other systems. The RRC layer also performs the reporting of the measurements from the UE to the network. Outer loop power control: the RRC layer controls setting of the target of the closed loop power control.

66

Broadband Wireless Mobile: 3G and Beyond

† Control of ciphering: the RRC layer provides procedures for setting of ciphering (on/off) between the UE and UTRAN. † Slow DCA: allocation of preferred radio resources based on long-term decision criteria. It is applicable only in TDD mode. † Broadcast of ODMA relay node neighbour information: the RRC layer performs probe information broadcasting to allow ODMA routing information to be collected. † Maintenance of ODMA relay nodes neighbour lists and gradient information: the ODMA relay node neighbour lists and their respective gradient information will be maintained by the RRC. † Maintenance of number of ODMA relay node neighbours: the RRC will adjust the broadcast powers used for probing messages to maintain the desired number of neighbours. † Establishment, maintenance and release of a route between ODMA relay nodes: the establishment of an ODMA route and RRC connection based upon the routing algorithm. † Interworking between the gateway ODMA relay node and the UTRAN: the RRC layer will control the interworking with the standard TDD or FDD communication link between the gateway ODMA relay node and the UTRAN. † Paging/notification: the RRC layer can broadcast paging information from the network to selected UE. Higher layers on the network side can request paging and notification. The RRC layer can also initiate paging during an established RRC connection. † Initial cell selection and re-selection in idle mode: selection of the most suitable cell based on idle mode measurements and cell selection criteria. † Arbitration of radio resources on uplink DCH: this function controls the allocation of radio resources on uplink DCH on a fast basis, using a broadcast channel to send control information to all involved users. † RRC message integrity protection: this function adds a Message Authentication Code (MAC-I) to those RRC messages that are considered sensitive and/or contain sensitive information. † Timing advance (TDD mode): the RRC controls the operation of timing advance. It is applicable only in TDD mode.

Figure 2.22 shows the RRC model for the UE side. The functional entities of the RRC layer are described below: † Routing of higher layer messages to different Mobility Management/Connection Management (MM/CM) entities (UE side) or different core network domains (UTRAN side) is handled by the Routing Function Entity (RFE). † Broadcast functions are handled in the broadcast control function entity (BCFE). The BCFE is used to deliver the RRC services, which are required at the GC-SAP. † Paging of idle mode UE(s) is controlled by the paging and notification control function entity (PNFE). The PNFE is used to deliver the RRC services that are required at the NtSAP. † The Dedicated Control Function Entity (DCFE) handles all functions specific to one UE. The DCFE is used to deliver the RRC services that are required at the DC-SAP. † In TDD mode, the DCFE is assisted by the Shared Control Function Entity (SCFE) location in the C-RNC, which controls the allocation of the PDSCH and PUSCH.

UMTS Air Interface

67

Figure 2.22 RRC model for the UE side.3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

† The Transfer Mode Entity (TME) handles the mapping between the different entities inside the RRC layer and the SAPs provided by RLC.

RRC procedures Radio resource control functions [40] are accomplished by means of several procedures. They can be classified into the following: † † † † † †

RRC connection management procedures Radio bearer control procedures RRC connection mobility procedures Measurement procedures General procedures Generic actions on receipt of an information element.

68

Broadband Wireless Mobile: 3G and Beyond

Tables 2.10 2.15 list the procedures under each class. Table 2.10 RRC connection management procedures Broadcast of system information Paging RRC connection establishment RRC connection release RRC connection re establishment Transmission of UE capability information UE capability enquiry Initial Direct transfer

Downlink direct transfer Uplink direct transfer UE dedicated paging Security mode control Signalling flow release procedure Signalling connection release request procedure Counter check

Table 2.11 Radio bearer control procedures Radio bearer establishment Reconfiguration procedures Radio bearer release Transport channel reconfiguration Transport format combination control Physical channel reconfiguration Physical shared channel allocation (TDD only)

PUSCH capacity request (TDD only) Downlink outer loop control Uplink physical channel control (TDD only) Physical channel reconfiguration failure

Table 2.12 RRC connection mobility procedures Cell update URA update UTRAN mobility information Active set update Hard handover Channel allocation (TDD only)

Inter Inter Inter Inter

system handover to UTRAN system handover from UTRAN system cell reselection to UTRAN system cell reselection from UTRAN

Table 2.13 Measurement procedures Measurement control

Measurement report

69

UMTS Air Interface Table 2.14

General procedures

Selection of initial UE identity Actions when entering idle mode from connected mode Open loop power control upon establishment of DPCCH Physical channel establishment criteria Detection of out of service area Radio link failure criteria Generic state transition rules depending on received information elements Open loop power control

Table 2.15

Detection of in service area Hyper frame numbers START Integrity protection Measurement occasion calculation Establishment of access service classes Mapping of access classes to access service classes PLMN type selection CFN calculation

Generic actions on receipt of an information element

CN information elements UTRAN mobility information elements UE information elements Radio bearer information elements

Transport channel information elements Physical channel information elements Measurement information elements Other information elements

Interactions between RRC and lower layers Besides the immediate lower sub-layer (RLC), RRC interacts with the other lower layers by controlling the allocation of radio resources and allowing MAC to arbitrate between users and radio bearers. RRC uses the measurements done by the lower layers to determine which radio resources are available. Figure 2.23 shows SAPs between radio resource control and lower layers. User equipment identification during a RRC connectionA Radio Network Temporary Identity (RNTI) is used as a UE identifier when a RRC connection exists. There are two types of RNTIs depending on their use: S-RNTI is used within the serving RNC and C-RNTI is used within a cell controlled by a controlling RNC. Additionally there is an identifier for the current serving RNC denoted as a SRNC identifier. The combination of S-RNTI and SRNC is referred to as U-RNTI (UTRAN Radio Network Temporary Identity) and it is used in the radio interface for functions such as cell update or paging. There are two different levels of UE connection to UTRAN. They are listed below: † No signalling connection exists: the UE has no relation to UTRAN, only to CN. For data transfer, a signalling connection has to be established. † Signalling connection exists: there is a RRC connection between UE and UTRAN. The UE position can be known on different levels:

70

Broadband Wireless Mobile: 3G and Beyond

Figure 2.23 Interaction between RRC and lower layers. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

† SUBUTRAN Registration Area (URA) level: UE position is known on UTRAN registration area level. URA is a specified set of cell, which can be identified on the BCCH. † SUBCell level: UE position is known on cell level. Different channel types can be used for data transfer: common transport channels (RACH, FACH, CPCH, DSCH); dedicated transport channels (DCH).

2.3 CDMA2000 Air Interface The cdma2000 is a family of standards that defines the air interface, minimum performance and service supported by the standard. It meets all the requirements for the next generation evolution of the current TIA/EIA-95-B family of standards and as part of the ‘family of systems’ concept designed by ITU to be able to connect different radio transmission modules to the same core network equipment. The general model for the radio interface in the case of UMTS is supported by a layered structure with a protocol stack and information flow clearly defined for each of the layer interfaces. The layers defined are: Physical Layer (L1), Data Link Layer (L2) which is further subdivided into MAC Sublayer and LAC Sublayer and Upper Layers. Services provided, signalling, data and voice, are logically resident in the upper layers; those services should be translated to logical channels containing control and user information.

UMTS Air Interface

71

The upper layer controls the messaging interchange given by the protocol communication process accordingly to an established procedure. The upper layer makes use of the services provided by L2. Those services are the correct transmission and reception of signalling messages, and are supported partly in MAC and LAC sublayer. LAC implements a data link protocol that transports and delivers correctly the signalling messages generated in the upper layer. In the same way, LAC makes use of the services offered from MAC layer, i.e. the transportation of LAC protocol units through the medium access protocol. MAC is the functional entity that controls the access of L1 resources to services resident in upper layer. At the last step L1 supports the part of the communication protocol between the mobile station and the base station that is responsible for the transmission and reception of data. MAC layer provides the physical layer with frames and L1 should process it accordingly in order to support the data transmission and reception. All the protocol communication can be summarized as an interchange of information between layers supported by the services provided by the lower layers.

2.3.1 Layer 1 The cdma2000 radio interface [44] is a wideband, spread spectrum interface that uses CDMA technology. It satisfies the ITU requirements for the indoor office, indoor to outdoor pedestrian, and vehicular environments. The key design characteristics include W-CDMA radio interface offering significant advances to increase performance and capacity; data rates form 1.2 kbps to greater than 2 Mbps; support a wide range of RF channel bandwidths, advanced Medium Control Access (MAC); turbo codes for higher transmission rates and increased capacity; support for forward radio interface transmit diversity, etc. The cdma2000 standard matches the ITU requirements for the IMT200 CDMA MultiCarrier Mode (MC-CDMA). This mode uses N (N {1,3}) adjacent 1.2288 Mcps DirectSequence (DS) spread RF carriers for the forward (downlink) channel and a single DS spread RF carriers with spreading rates (SR) SR1 1.2288 Mcps SR3 3.6864 Mcps for the reverse (uplink). The system may work then for two different spreading rates SR1 and SR2 and different frequency bands specified by the band class parameter. The Mobile Station (MS) supports two spreading rates based on spreading a single carrier with different SR. On the other hand the Base Station (BS) supports the two spreading rates based on multiple carriers with just one 1.228 Mcps spreading rate; for SR1 just one carrier is used, for SR3 three carriers are used. From the point of view of how the radio resource is accessed for forward and reverse channels the system supports frequency division like in FDD mode in UMTS. A physical layer description should differentiate the forward link, the reverse link and the radio resource function. There are several characteristics within the forward link physical layer that include: † Common pilot: the system provides a common code multiplexed pilot for all users. This pilot channel is shared by all traffic channels and thus it provides an efficient utilization of resources. It is used with the purpose of estimating channel gain and phase, detecting multipath rays and for cell acquisition and handoff. † Auxiliary pilots: certain applications such as antenna arrays and antenna transmit diversity require a separate pilot for channel estimation and phase tracking

72

Broadband Wireless Mobile: 3G and Beyond

† Independent data channels: the systems provides two types of forward link data channels (fundamental and supplemental). Each can be adapted to a particular kind of service.

The reverse link provides a continuous waveform for all data rates. This includes continuous pilot and continuous data channel waveforms. It also permits a range increase at lower transmission rates, and the interleaving to be performed over the entire frame (so the interleaving achieves the full benefit of the frame time diversity). The base station uses the pilot for multipath searches, tracking, coherent demodulation and to measure the quality of the link control purposes. It uses separate orthogonal channels for the pilot and each of the data channels. There are several characteristics within the reverse link physical layer and includes: † Orthogonal channels provided using different length Walsh sequences: the system uses orthogonal channels for the pilot and other physical data channels. † Rate matching: through puncturing, symbol repetition and sequence repetition. † Low spectral sidelobes: the sidelobes generated by the non-ideal mobile power amplifiers are controlled by splitting the physical channel between the in-phase and quadrature data channels. † Power control. † Separate dedicated control channel is a separate, low rate, low power, continuous orthogonal dedicated control channel.

The radio resource function include code planning, system acquisition, handoff procedures, reverse common access procedures. Logical and physical channelsA logical channel is a communication path described in terms of the intended use (for a single target or for multiple targets), the transferred data (signalling or user traffic data) and the direction of the transfer (forward or reverse). Given these three criteria logical channels are named with three lower case letters (Table 2.16) followed by ‘ch’ (channel). Table 2.16

Naming conventions for logical channels

First letter

Second letter

Third letter

f ¼ forward r ¼ reverse

d ¼ dedicated c ¼ common

t ¼ traffic s ¼ signalling

But not all combinations of the three letters for naming logical channels are possible, i.e. a common traffic channel does not exist. The rest of the combinations f-csch, f-dsch, f-dtch, rcsch, r-dsch and r-dtch are possible, supporting the following functionalities needed for the provision of services: synchronization, broadcast, general signalling, access and dedicated signalling. A physical channel is the mapping of a logical channel to a particular radio resource; the information is protected and set up to send through the radio channel. The same way that logical channels are classified for the intended use, physical channels can be broadly categorized into two basic classes:

73

UMTS Air Interface

† Dedicated channels: carry information in a dedicated point-to-point manner between the base station and a single mobile station. † Common channels: carry information in a shared access, point-to-multipoint manner between the base station and multiple mobile stations.

A logical channel can be mapped to one or more physical channels. A logical channel may have permanent and exclusive use of a physical channel (e.g. the synchronization channel), may have temporary, but exclusive while the time that is being used, of a physical channel or may share the physical channel with other logical channels. Permanent mapping is clearly defined, for example the f-csch carrying synchronization information is permanently mapped to the Forward Sync Channel (F-SYNCH) and the f-csch carrying broadcast information is permanently mapped to the Forward Broadcast Control Channel (F-BCCH). A Logical-to-Physical Mapping (LPM) table defines temporary mappings. LPM is a nonnegotiable service configuration parameter where each entry to the table consists of a service reference identifier, the logical resource, the physical resource, the forward/reverse flag and the priority. Figures 2.24 and 2.25 show the mapping from logical channels to physical channels for both forward and reverse connections. All the possible physical channels are defined in Table 2.17.

Figure 2.24

Mapping of forward logical channels to physical channels.

Figure 2.25

Mapping of reverse logical channels to physical channels.

74

Broadband Wireless Mobile: 3G and Beyond Table 2.17

Physical channels in cdma2000

Channel name

Meaning

F/R FCH F/R DCCH F/R SCCH F/R SCH F PCH F QPCH R ACH F/R CCCH F/R PICH F APICH F TDPICH F ATDPICH F SYNCH F CPCCH F CACH R EACH F BCCH

Forward/reverse fundamental channel Forward/reverse dedicated control channel Forward/reverse supplemental code channel Forward/reverse supplemental channel Paging channel Quick paging channel Access channel Forward/reverse common control channel Forward/reverse pilot channel Dedicated auxiliary pilot channel Transmit diversity pilot channel Auxiliary transmit diversity pilot channel Sync channel Common power control channel Common assignment channel Enhanced access channel Broadcast control channel

Mapping of logical dedicated channels All the physical channels associated with the logical dedicated traffic and signalling channel, in both directions forward or reverse, share several common characteristics. These characteristics are grouped and named under radio configuration. There are six radio configurations defined for the reverse link and nine for the forward; each of them define the associated spreading rate SR1 or SR3, the data rates, coding rates and modulation for the physical channels mapped with this radio configuration. The mapping of r-dtch and r-dsch for radio configurations 1 and 2 is composed of a R-FCH and up to seven R-SCCH. For radio configurations 3 6, is composed of a R-FCH, a R-DCCH and up to two R-SCH. The f-dtch and f-dsch mapping, for radio configurations 1 and 2 is composed of one F-FCH, and up to seven F-SCCH. For radio configurations 3 9 is composed of a F-FCH, a F-DCCH and up to two F-SCH. The F-FCH carries a combination of higher level data and power control information; the associated reverse channel carries higher level data and control information. Transmission rates for the forward fundamental channel are dependent on the radio configuration, the system can work with flexible (radio configurations 3 9) or non-flexible rates (radio configurations 1 9).In the reverse channel both working modes are possible too: flexible mode for radio configurations 3 6 and non-flexible modes for all radio configurations. For both forward and reverse communications, flexible mode provides a range of variations from 850 bps to 14.48 kbps; non-flexible mode varies in the range 1.2 kbps to 14.48 kbps. From the mapping for different radio configurations the supplemental channel and the supplemental code channel are not used simultaneously, but both of them allow, in the

UMTS Air Interface

75

reverse and forward directions, the support of higher data rate services. F-SCH and R-SCH have a flexible and non-flexible rate working mode. The range of variation for both modes depends on frame duration and the radio configuration mode. The frame length belongs to the set {20 ms, 40 ms, 80 ms}. The flexible rate has subintervals for range variation defined between 921.6 kbps and 750 bps. For non-flexible mode rates are defined in the range 1036.8 1.2 kbps. F-SCCH and R-SCCH work in a non-flexible mode with frame duration of 20 ms. the rates depend on the radio configurations: radio configuration 1 works at 9.6 kbps and radio configuration 2 at 14.4 kbps. For F-DCCH transports higher level data, control information and power control information, the associated reverse channel carries the same except for the power control information. The dedicated control channel can work at non-flexible rates: 9.6 kbps with frame duration of 5 ms or 20 ms and 14.48 kbps with frame duration of 20 ms. For the flexible rates the frame duration is fixed: 20 ms, and the range of variation depends on the radio configuration but goes from 1.05 kbps to 144 kbps. Mapping of logical common channels The remaining physical channels are related to the f-csch and the r-csch. In the forward communication there are two important channels with a permanent assignation to the radio resource, the F-SYNCH and the F-BCCH. The F-SYNCH transports the synchronization message to the mobile station. The synchronization message is a coded, interleaved, spread and modulated spread spectrum signal. The Sync channel operates in all the coverage area of the base station transmitting the channel. It has a fix rate of 1.2 kbps. F-BCCH is used for the transmission of general-purpose control information and also paging. F-BCCH operates within the coverage area of the base station. The possible transmission rates are 19.2 kbps, 9.6 kbps and 4.8 kbps. Pilot channels are unmodulated direct-sequence spread spectrum signal transmitted by a BS or MS. A pilot channel provides a phase reference for coherent demodulation and may be used for power measurements for handoff. In the forward link the pilot channels support partly the synchronization process of the mobile station too. The F-PICH is transmitted all the time with each active forward-dedicated channel. If the base station is using transmit diversity the F-TDPICH is used. The paging channel is defined exclusively in the forward communication. It applies only to the SR1. The usual operations are performed over this channel: coding, interleaving, spreading and modulation. The paging channel is used to transmit overhead information and messages to specific mobile stations. The frame duration is fixed and the transmission rates are 4.8 kbps and 9.6 kbps. Unlike the paging channel, F-QPCH is not encoded. It is used to inform mobile stations in idle state whether or not they should receive the F-CCCH or the F-PCH in the next time slot. The common control channel is used for the transmission of user-specific and control information. In the reverse link it can be used for one or several MS for transmitting user and signalling information when the reverse traffic channels are not in use. The transmission is in time slots defined by the base station. In the forward link the common control channel is used to transmit messages to MS. The transmission rates for both, reverse and forward channel are 9.6 kbps, 19.2 kbps and 38.4 kbps in radio frames of 20 ms, 10 ms or 5 ms. F-CPCCH is used by the base station for transmitting information for the power control of multiple R-CCCH and R-EACH. F-CACH supports the transmission of random access pack-

76

Broadband Wireless Mobile: 3G and Beyond

ets on the reverse link for R-CCCH, providing channel assignments in the reverse link. This channel works at a fixed rate, 9.6 kbps with a frame duration of 5 ms. The R-ACH is used by the mobile station to initiate communication with the BS, to respond to paging messages and to register. Its main physical characteristic is that it is a slotted random-access channel and is uniquely identified by long codes. R-EACH has the additional characteristic that can be used in three possible modes: basic access mode, power controlled access mode and reservation access mode. The basic mode is very similar to the R-ACH working procedure; the power controlled mode uses a close loop power control and the reservation mode transmits the enhanced access data in the R-CCCH with close loop power control. Coding of physical channels The coding schemes proposed for forward error correction are convolutional or turbo encoding. Which one is used, and which coding rates are applied, depends on the type of channel that needs to be protected against errors and, if it applies, the Radio Configuration (RC) that is being used in the communication (Tables 2.18 and 2.19). Convolutional encoding is applied with coding rates 1/2, 1/3, 1/4 and 1/6 and memory order 8. The generator polynomials are given in Table 2.20. The turbo coding scheme is based on the parallel concatenation of two identical recursive systematic convolutional encoders of rate 1/3. The memory order is 3 for each of the convolutional encoder and the generator polynomials are in octal: 13 for the recursive bit, g1 15 and g2 17. Puncturing of the output sequence is performed in order to obtain the respective coding rates. The second encoder input is an interleaved version of the input to the first one. The turbo interleaver is based on a block interleaver where inter-row and intra-row permutations are performed. Spreading and modulation In the reverse communication the orthogonality between simultaneously transmitted symbols is assured with a 64-ary orthogonal modulation or a orthogonal spreading based on Walsh sequences. Both procedures already spread the signal though not in all cases to the required SR1 or SR3. For SR1 the final spread is accomplished by a long code with period 2 42 1 chips. Different sequences can be generated by masking the output sequence with a 42-bit mask. The mask identifies the particular channel number that is being spread. For SR3 the long sequence is generated by multiplexing three long sequences generated by delaying and operating over the 1.2288 Mcps sequence. Additionally the reverse physical channel may be quadrature spread with a sequence of period 2 15 for SR1 or period 2 20 1 for SR3.

2.3.2 Layer 2 Data Link Layer (Layer 2) in cdma2000 air interface is divided into two sub-layers, according to the general structure specified by ITU for IMT-2000 systems: Medium Access Control (MAC) and Link Access Control (LAC). The medium access control sub-layer interacts directly with Layer 1. Applications and upper layer protocols corresponding to OSI layers

Convolutional

Convolutional

Convolutional Convolutional or turbo; (N $ 360)

Convolutional – Convolutional – Convolutional – Convolutional Convolutional

F-FCH

F-DCCH

F-SCCH F-SCH

F-PCH F-QPCH F-CCCH F-PICH F-SYNCH F-CPCCH F-CACH F-BCCH

1/2 – 1/2 or 1/4 – 1/2 – 1/2 or 1/4 1/2 or 1/4

1/2 (RC 1 and 2) 1/2 (RC 4); 1/4 (RC 3 and 5)

1/2 (RC 1, 2 and 4); 1/4 (RC 3 and 5)

1/2 (RC 4); 1/4 (RC 3 and 5)

SR1 Forward error correction Rate

Channel

Convolutional or turbo; (N $ 360) – – Convolutional – Convolutional – Convolutional Convolutional

– Convolutional

Convolutional

Convolutional

– – 1/3 – 1/2 – 1/3 1/3

1/3 (RC 7); 1/4 (RC 8); 1/2 (RC 9);

1/6 (RC 6); 1/3 (RC 7); 1/4 (RC 8, 20 ms); 1/3 (RC 8, 5 ms); 1/2 (RC 9, 20 ms); 1/3 (RC 9, 5 ms) 1/6 (RC 6); 1/3 (RC 7); 1/4 (RC 8, 20 ms); 1/3 (RC 8, 5 ms); 1/2 (RC 9, 20 ms); 1/3 (RC 9, 5 ms) – 1/6 (RC 6)

SR3 Forward error correction Rate

Table 2.18 Coding schemes for forward channels (radio configuration and coding rates)

UMTS Air Interface 77

Convolutional

Convolutional Convolutional Convolution; or turbo; N $ 360

Convolutional Convolutional – Convolutional

R-FCH

R-DCCH R-SCCH R-SCH

R-ACH R-CCCH R-PICH R-EACH

1/3 (RC 1); 1/2 (RC 2); 1/4 (RC 3 and 4) 1/4 1/3 (RC 1); 1/2 (RC 2) 1/4 (RC 3, N , 6120); 1/2 (RC 2, N ¼ 6120); 1/4 (RC 4) 1/3 1/4 – 1/4

SR1 Forward error correction Rate

Channel

– Convolutional – Convolutional

Convolutional – Convolution; or turbo; N $ 360

Convolutional

– 1/4 – 1/4

1/4 – 1/4 (RC 5, N , 6120); 1/3 (RC 5, N $ 6120); 1/4 (RC 6, N , 20712); 1/2 (RC 6, N ¼ 20712)

1/4

SR3 Forward error correction Rate

Coding schemes for reverse channels (radio configuration and number N of channel bits per frame)

Table 2.19

78 Broadband Wireless Mobile: 3G and Beyond

79

UMTS Air Interface

Table 2.20

Generator polynomials for convolutional encoding

Rate

g0 (oct)

g1 (oct)

g2 (oct)

g3 (oct)

g4 (oct)

g5 (oct)

1/6 1/4 1/3 1/2

457 765 557 753

755 671 663 561

551 513 711

637 473

625

727

3 7 utilize the services provided by link access control. The situation of these two sub-layers in the context of the air interface protocol stack is shown in Figure 2.26. Medium access control (MAC)The cdma2000 MAC sub-layer [45] provides two important functions whose situation in the layer structure is shown in Figure 2.26: † Multiplexing and QoS control by prioritization of the access requests. † Best effort delivery via a Radio Link Protocol (RLP).

In addition, a Signalling Radio Burst Protocol (SRBP) provides a connectionless service for signalling messages. MAC sub-layers and their interaction with the physical layer are depicted in Figure 2.27. Multiplexing and QoS sub-layer is subdivided into a multiplex sublayer and a common channel multiplex sub-layer for transmission and reception of dedicated and common information and signalling, respectively. Multiplex and common channel multiplex sub-layers The multiplex sub-layer has both a transmitting and receiving function. The transmitting function combines information from upper layers and it conveys it to the physical layer for

Figure 2.26 MAC and LAC in cdma2000 layer structure.

80

Broadband Wireless Mobile: 3G and Beyond

Figure 2.27 MAC sub layers and interaction with physical layer in cdma2000.

transmission. The receiving function separates the information coming from the physical layer and directs it to the proper upper layer entity. Since cdma2000 and TIA/EIA-95-B must be interoperable, the multiplex sub-layer is specified to work in two different modes: † Mode A is used when operating with a radio configuration less than or equal to 2. † Mode B is used when operating with a radio configuration greater than 2.

Information bits are exchanged between a connected service or a logical channel and the multiplex sub-layer in a unit called a data block. The multiplex sub-layer multiplexes one or more data blocks into a MuxPDU (multiplex protocol data unit) and combines one or more MuxPDUs into a physical layer SDU (service data unit) for transmission. The receiving function accepts physical layer SDU, divides it into one or more MuxPDUs, demultiplexes the data blocks in each MuxPDU and delivers the information to the appropriate service. Several types of MuxPDUs are defined to allow multiplexing the different dedicated physical channels (F/R-FCH, F/R-DCCH, F/R-SCCH and F/R-SCH) in 20 ms and 5 ms data blocks.

81

UMTS Air Interface

Figure 2.28

Multiplex sub layer transmitting function example.

Figure 2.28 shows an example of multiplexing of data blocks into different types of MuxPDUs. Mapping between physical and logical channels is specified in a Logical to Physical Mapping (LPM). The common part of the multiplex sub-layer establishes the common physical channels. Access channel procedures The entire process of sending one Layer 2 encapsulated PDU and receiving (or failing to receive) an acknowledgement for the PDU is called an access attempt. One access attempt consists of one or more access sub-attempts. Each transmission in the access sub-attempt is called an access probe and consists of an R-ACH preamble and an R-ACH message. Within an access sub-attempt, access probes are grouped into access probe sequences. Access probes in every sequence are transmitted with increasing power level. The R-ACH used for each access probe sequence is chosen pseudo-randomly from among all the R-ACHs associated with the current F-PCH. The timing of access probes and access probe sequences is expressed in terms of R-ACH slots. Transmission of an access probe begins at the start of one pseudo-randomly selected RACH slot. For every access probe sequence a backoff delay is generated pseudo-randomly. An additional delay is imposed by the use of a random persistence test. Upon reception of an acknowledgement from the base station, the access is terminated. When the mobile station fails to receive its acknowledgement on the F-CACH within a time limit, the SRBP entity sends a failure indication to the LAC sub-layer. R-EACH channels and enhanced access probes can be used to obtain enhanced access. There are two different access modes that can be used when transmitting messages using the enhanced access procedures:

82

Broadband Wireless Mobile: 3G and Beyond

† basic access mode † reservation access mode.

The SRBP entity uses an algorithm to select an access mode based on the length of the Layer 2 PDU and configuration parameters from the base station. Link access control (LAC) The LAC sub-layer [46] implements a data link protocol that provides for the correct transport and delivery of signalling messages generated by Layer 3. SDUs are passed between Layer 3 and the LAC sub-layer and they are encapsulated into LAC PDUs, which are subject to segmentation and reassembly and are transferred to the MAC sublayer. SDUs and PDUs are processed and transferred along functional paths, the logical channels, without the need for the upper layers to be aware of the radio characteristics of the physical channels. As a generated or received data unit traverses the LAC protocol stack, it is processed by various protocol sub-layers in sequence, each of which processes only specific fields of the data unit. The general processing of data units by the LAC and its sub-layers is shown in Figure 2.29.

Figure 2.29

LAC data unit processing.

83

UMTS Air Interface

Figure 2.30

cdma200 forward logical channels.

Logical channels Logical channels are classified depending on the type and direction of the information carried as: † † † † †

synchronization broadcast general signalling access dedicated signalling.

Figures 2.30 and 2.31 show all cdma2000 logical channels in forward and reverse directions. Functions and protocol architecture of the LAC sub-layer The LAC sub-layer performs the following functions: † reliable delivery of SDUs to the peer Layer 3 using ARQ techniques; † assembling and validating appropriate PDUs to carry SDUs.

84

Broadband Wireless Mobile: 3G and Beyond

Figure 2.31

† † † †

cdma200 reverse logical channels.

segmentation of encapsulated PDUs into fragments of suitable size; reassembly of encapsulated PDU fragments; access control though authentication; address control to ensure PDU delivery to particular mobile stations.

These functions are carried out by five LAC sub-layers: ARQ, utility, SAR, authentication and addressing sub-layers. ARQ model The ARQ sub-layer for each logical channel is split between the transmitter and the receiver side and provides two major types of service to Layer 3: † Assured delivery service: PDUs are repeatedly sent at fixed intervals until an acknowledgement is received or a specific number of retransmissions is reached, in which case the logical channel is dropped. † Unassured delivery service: PDUs are not acknowledged and there is no guarantee of reception.

UMTS Air Interface

85

When requesting transmission, Layer 3 specifies which type of service delivery is to be used. Layer 3 can also request the LAC to perform a reset of the ARQ procedures.

2.3.3 Layer 3 Layer 3 [47] generates Layer 3 PDUs and passes them to lower layers for proper encapsulation and transmission. On the receiving end, PDUs are decapsulated and sent to Layer 3 through SDUs from lower layers. The interface between Layer 3 and Layer 2 is a Service Access Point (SAP), at which Service Data Units (SDUs) and control information in the form of Message Control and Status Blocks (MCSB) are interchanged. In the data plane, Layer 3 originates and terminates signalling data units according to the semantic and timing of the communication protocol between the base station and the mobile station. From a semantic point of view, the signalling data units are referred to as ‘messages’ (or ‘orders’). From a protocol point of view, the signalling data units are PDUs. Message control and status blocks (MCSB) The MCSB is a parameter block containing relevant information about individual Layer 3 messages as well as instructions on how these messages may be handled and processed in transmission or reception by Layer 2. MSCB must contain information such as: † indication whether the message is generated in response to a previously received message; † length of the message; † unique instance identifier associated with the message, which enables identification of a message for notifications of delivery/non-delivery or recovery procedures; † indication whether the message should be acknowledged at Layer 2 (i.e., delivered in assured or unassured mode); † indication whether notification of delivery is required; † identity of the addressee for the message; † indication whether the PDU delivered to Layer 3 is a duplicate; † data needed by the authentication procedures; † relevant PDU classification, when processing at Layer 2 is sensitive to the kind of PDU being transferred; † encryption status of the logical channel; † CDMA system time corresponding to the frame in which the first or last bit of a message was received; † transmission instructions for Layer 2, such as an instruction to send a message with a certain priority; † abnormal conditions indications from Layer 2.

Interaction with Layer 2 for PDU transmission and reception Layer 3 employs the services offered at the interface with Layer 2 to transfer PDUs to and from the Layer 3 entity. When requesting the transmission of a PDU, Layer 3 will typically specify whether the transfer will be performed in assured mode or in unassured mode (for example, by setting the proper parameters in the MCSB argument of the data request primi-

86

Broadband Wireless Mobile: 3G and Beyond

tive, as explained before). For transmission in assured mode, Layer 3 may specify if confirmation of delivery of the PDU is required. Layer 2 guarantees that an assured mode PDU received from the transmitting Layer 3 entity is delivered to the receiving Layer 3 entity. Each assured mode PDU is delivered to the receiving Layer 3 entity only once and without errors. Additionally, if the transmitting Layer 3 entity requests confirmation of delivery of an assured mode PDU, Layer 2 will send an indication to the transmitting Layer 3 entity when Layer 2 receives an acknowledgement for that PDU. If Layer 2 is not able to deliver an assured mode PDU, it sends an indication of the failure to Layer 3, which can then take corrective action. Layer 2 does not guarantee that an unassured mode PDU received from the transmitting Layer 3 entity is delivered to the receiving Layer 3 entity. Thus, Layer 2 acknowledgements may not be required for unassured mode PDUs. To increase the probability of delivery of unassured mode PDUs, Layer 3 may request Layer 2 to send the PDUs many times and rely on the duplicate detection capabilities of the receiver to achieve uniqueness of delivery. Layer 3 can also request Layer 2 to perform a reset of the Layer 2 ARQ procedures. Security and identification Mobile stations operating in the CDMA mode are identified by an International Mobile Station Identity (IMSI). Mobile stations shall have two different identifiers, IMSI T (true IMSI) and IMSI M (MIN-based IMSI). IMSI consists of up to 15 numerical characters (0 9). The first three digits of the IMSI are the Mobile Country Code (MCC) and the remaining digits are the National Mobile Station Identity (NMSI), consisting of the Mobile Network Code (MNC) and the Mobile Station Identification Number (MSIN). Two classes of IMSI are defined depending on length: † Class 0 IMSI: an IMSI that is 15 digits in length IMSI (the NMSI is 12 digits in length) † Class 1 IMSI: an IMSI that is less than 15 digits in length is called a class 1 IMSI (the NMSI is less than 12 digits in length).

IMSI M is an IMSI that contains a MIN (Mobile Identification Number) in the lower ten digits of the NMSI. IMSI T is an IMSI that is not associated with the MIN assigned to the mobile station. When operating in the CDMA mode the mobile station shall set its operational IMSI value, IMSI O, to either the IMSI M or the IMSI T depending on the capabilities of the base station. Three more numbers are used for identification of the mobile stations: † Mobile Directory Number (MDN) associated with the mobile station through a service subscription. A MDN is not necessarily the same as the mobile station identification on the air interface, i.e. MIN, IMSI M or IMSI T. † Electronic Serial Number (ESN) a 32-bit binary number that uniquely identifies the mobile station to any wireless system. † Station Class Mark (SCM) specifying whether the mobile is Band Class 0 or Band Class 1, dual mode, capable of discontinuous transmission. † Temporary Mobile Station Identity (TMSI) is a temporary locally assigned number that the mobile station obtains when assigned by the base station and does not have any association with the mobile station’s IMSI, ESN, or directory number.

87

UMTS Air Interface

A TMSI zone is an arbitrary set of base stations for the administrative assignment of TMSIs, identified by a TMSI ZONE field. A TMSI CODE is uniquely assigned to a mobile station inside a TMSI zone and may be reused to identify a different mobile station in a different TMSI zone. The pair (TMSI ZONE, TMSI CODE) is a globally unique identity for the mobile station called the full TMSI. The TMSI CODE can be two, three, or four octets in length. The TMSI ZONE can range from 1 to 8 octets in length.

Figure 2.32

SSD generation and subsets.

Authentication M IMSI and T IMSI are used for the authentication process: M IMSI will be used if it is available, otherwise, T IMSI shall be used. A successful outcome of the authentication process occurs only when it can be demonstrated that the mobile station and base station possess identical sets of Shared Secret Data. SSD is a 128-bit quantity that is stored in semipermanent memory in the mobile station and is readily available to the base station. Figure 2.32 shows the process of SSD generation. SSD is partitioned into two distinct 64-bit subsets. Each subset is used to support a different process: SSD A is used to support the authentication procedures and SSD B is used to support voice privacy and message encryption. The SSD shall not be accessible to the user. Privacy and encryption Voice privacy is provided in the CDMA system by means of the private long code mask used for PN spreading. Voice privacy is provided on the traffic channels only. All calls are initiated using the public long code mask for PN spreading but the mobile station user may request voice privacy during or after call setup. Information privacy is assured by using an encryption algorithm. In an effort to enhance the authentication process and to protect sensitive subscriber information, a method is provided to encrypt selected f-dsch, r-dsch, f-csch or r-csch Layer 3 signalling PDUs.

88

Broadband Wireless Mobile: 3G and Beyond

Figure 2.33

CRC generation.

Before going through the encryption algorithm, the sender of the message shall append an 8-bit CRC to the end of the Layer 3 PDU. The generator polynomials for the 8-bit CRC field is gðxÞ x8 1 x7 1 x4 1 x3 1 x 1 1. CRC computation is shown in Figure 2.33. Layer 3 processing states Mobile station Layer 3 processing consists of the following states that are illustrated in Figure 2.34:

Figure 2.34

Mobile station Layer 3 processing states.

89

UMTS Air Interface

Figure 2.35

Mobile station initialization state.

† Mobile station initialization state: in this state, the mobile station selects and acquires a system. † Mobile station idle state: in this state, the mobile station monitors messages on the f-csch. † System access state: in this state, the mobile station sends messages to the base station on the r-csch and receives messages from the base station on the f-csch. † Mobile station control on the traffic channel state: in this state, the mobile station communicates with the base station using the f/r-dsch and f/r-dtch.

After power is applied to the mobile station, it shall enter the system determination substate of the mobile station initialization state with a power-up indication. Mobile station initialization state is further detailed in Figure 2.35. Handoff procedures The mobile station supports the following three handoff procedures while in the mobile station control on the traffic channel state: † Soft handoff: a handoff in which the mobile station commences communications with a new base station without interrupting communications with the old base station. Soft handoff can only be used between CDMA channels having identical frequency assignments.

90

Broadband Wireless Mobile: 3G and Beyond

† CDMA-to-CDMA hard handoff: a handoff in which the mobile station is transitioned between disjoint sets of base stations, different band classes, different frequency assignments, or different frame offsets. † CDMA-to-analogue handoff: a handoff in which the mobile station is directed from a CDMA traffic channel to an analogue voice channel.

2.4 Compatibility Issues Compatibility is a quite generic term to address the cooperation between two networks. Under this term, several different meanings are understood depending on the speaker or context. For example when a terminal manufacturer addresses compatibility, usually it means dual terminals being able to connect two different standards of network. When a network manufacturer uses the term compatibility, it usually means interconnectivity between two networks. Finally, when a operator thinks of compatibility it might mean co-existence of two standards without interfering with each other, or the reachability (and charging) of clients even if they need to change terminals. In the context of this book, by ‘compatibility’ we will understand all the issues that may concern the different players in the 3G mobile networks: final users, manufactures, operators, and administrations. In particular, the following issues are addressed: † Coexistence of any pair of systems, assuming they are not coordinated and no cooperation (roaming) is intended. On this point, the possible interference to a system caused by different standards will be analysed. Two analysis are required: † SUBpossibility of co-locating base-stations of 3G and 2G in the same site. Because 2G and 3G portions of spectrum are well separated, interference between them is not an issue, except for the case of placing a Node-B of 3G in the same location than a BTS of 2G; because of the adversity of this scenario, spurious isolation mask has to be checked; † SUBpossible interferences occurred when coexisting different standards or modes in 3G in the same geographical area; coexistence between modes FDD and TDD of UMTS deserves special attention. † Coexistence of unlicensed TDD UMTS networks in the same geographical area. In this particular case, the relevant administrations will not perform the frequency planning to guarantee coexistence. Thus inter-network interference may occur. † Cooperation between two different systems. Usually known as roaming, in this point the possibility of reaching a user who is ‘camped’ in a network under a different standard than the home network. Here the use of a different terminal is contemplated, but the issue is the possibility of reaching a user (defined by a phone number or IP address) in a different network. † Seamless handover. This is the most demanding point. When speaking of compatibility, the highest objective is to allow the users to switch from one to another network while in dedicated mode, without the user noting any disturbance.

A great amount of research is being conducted on compatibility between 3G mobile communications systems and digital broadcasting systems. The great potential for new applications of combining digital broadcasting networks and bi-directional mobile systems is pushing R&D in this direction. Examples of these applications are interactive broadcasting

91

UMTS Air Interface

or customer relationships in e-commerce. For the sake of conciseness, this broadest sense of compatibility is not addressed here. Compatibility issues between 3G mobile networks and short-range new systems as WLAN or Bluetooth are not addressed herein either, although we agree with the general opinion on their great commercial possibilities.

2.4.1 3GPP-3G First thing in exploring compatibility between UMTS FDD, TDD and CDMA2000 is to identify the spectral bands where they will be deployed. Figure 2.36 shows these bands. Looking at the above band allocation, it can be observed that seamless handover between CDMA2000 (PCS band in USA) and UMTS will not be simple. A user able to operate under UMTS and CDMA2000 with only one piece of hardware will require dual terminals. These dual terminals will have to be able to receive two different bands and two different duplex frequencies. Complexity of such dual terminals would be close to the complexity of two separate terminals. A second observation of Figure 2.36 indicates possible interferences between modes TDD and FDD of UMTS. Observe that the lower TDD band is contiguous to the FDD lower band. Co-locating base stations with such close frequencies in the same site may create harmful interferences.

Figure 2.36

3G spectrum bands in several regions.

There are also some other obvious considerations about compatibility. Compatibility may depend on the type of services being provided. For example, doing handover for different types of services (conversational, streaming, interactive, or background) will impose different performances to terminals and networks, and therefore depending on the type of service the compatibility may be compromised. Handover of background services may be seen as roaming rather than a handover; however handover of a conversational service requires swift signalling.

92

Broadband Wireless Mobile: 3G and Beyond

Coexistence or interferences Coexistence UMTS FDD-TDD The case of coexistence of FDD and TDD modes in geographically close areas deserves special attention. The lower part of the TDD spectrum is contiguous to the lower part of the FDD spectrum. Although the two parts of the spectrum are disjoined, they come together at 1920 MHz. It is well known that coexistence of FDD and TDD systems in general is somehow problematic. Frequency planning to avoid interference between two FDD systems is simpler because, among other reasons, uplink and downlink are clearly separated. This is not the case when a TDD system is deployed close to a FDD one. Shelf-planning techniques, as Dynamic Channel Allocations (DCA), which are very efficient to cope with TDD-TDD interference are not applicable for the TDD-FDD interference case. There are several sources of interference when having a TDD UMTS network deployed close to a FDD one: † TDD Node-B interference with FDD Node-B. The lower part of the FDD spectrum is dedicated to uplink, but the lower part of TDD spectrum is utilized for both uplink and downlink. Therefore both terminals and Node-B may interfere with FDD Node-B receivers nearby. Because of TDD Node-B interference into FDD Node-B, base stations of both modes should not be co-located on the same site when serving macro-cells. Node-B’s serving macro-cells transmit a high power level, and technology does not allow enough filtering of TDD interference. Co-location of both Node-B’s is possible in micro-cell environments. † TDD terminals interfere with FDD Node-B. This interference may be partially obviated if locating FDD Node-B antennas in locations where no terminals can reach close, i.e. building roofs or other inaccessible sites. † TDD terminals interfere with FDD terminals.

Coexistence between UMTS and CDMA2000 UMTS and CDMA2000 will be deployed in different regions; therefore the interference between them is not an issue. In a future scenario where regions deploy both systems careful frequency planning will be required. However, planning to coordinate UMTS-FDD and CDMA2000 will be similar to that required to separate two operators with the same standard. Coordination of UMTS-TDD and CDMA2000 will mean the same problems as coordination between FDD and TDD. Cooperation or roaming Both 3GPP and 3GPP2 standards ensure the interconnectivity of their core networks. Therefore, in principle, it will be possible to do ‘USIM-Roaming’ between CDMA2000 and UMTS networks. The term ‘USIM-Roaming’ means that users (defined by their unique international number or IP address) will be reachable when roaming into other networks, although they have to use a different terminal that meets the visited network standard. In any case, although the possibility is open, it will be up to the manufacturers and operators to implement such interconnectivity and to sign the corresponding inter-operators agreements.

UMTS Air Interface

93

Seamless handover The seamless handover between different 3G standards/modes will be possible only between UMTS FDD and TDD. A UMTS terminal operating in TDD mode can easily monitor the received level from the neighbouring FDD Node-B, because they continuously transmit the Pilot Channel (CPICH). A terminal operating in FDD mode can monitor the CPICH of TDD neighbouring cells which is transmitted twice every 10 ms. However, in order to have some idle time to do neighbouring cell measurements, the FDD link has to be in compressed mode, i.e. discontinuous transmission, with some idle time slots. Handover signalling in both modes are the same, therefore once the measurements are reported to the RNC rest of handover process is simple and perfectly implementable by networks. Note that handover between FDD and TDD will always be a hard-handover.

2.4.2 3G-2G The smooth transition from 2G to 3G mobile communications has been recognized as one of the main items relevant to the successful penetration of 3G in the worldwide market. Both proposed 3G technologies (UMTS and cdma2000) take into account this fact by ensuring interoperability and (in cdma2000) backwards compatibility with its previous 2G counterparts (GSM and IS-95). UMTS-GSM interoperability We have divided the different items involving UMTS-GSM interoperability into two nonexclusive areas. In some way this division is artificial, but we feel this is the correct way to address the different problems involved in this somewhat complex issue. User point of view In the evolution of the standards of UMTS, it had been proposed that the TDD mode of UMTS was designed with the GSM standards in view. The first TDD proposals ensured a mode of operation that was very similar to that of GSM (e.g. number and time period of slots). Designs oriented to take advantage of a common clock rates both for GSM and UMTS were advanced. This move was mainly oriented to reuse much of the hardware residing in the UE and the Node B for both modes. This initiative has finally come to nothing as the standards were modified in order to cope with other problems and specifications. The main consequence of the above is that for a UE to be usable in both UMTS and GSM environments, the hardware should be duplicated and superficial interoperation between both modules should be allowed. In order to minimize power consumption, when one of the modes (GSM or UMTS) was either in idle or connected mode, the other mode should be powered down. If a handover were necessary then it should be powered up and synchronized with the target BS or Node B. The increment in power consumption with respect to a GSM MS and a UMTS UE at the same bit rate should be negative as the better spectral efficiency of CDMA and the use of power-saving features as the gating mode, should give a gain in life time of the battery. As the first UE that will be in the market will not exceed 384 kbps, the life time of the battery should be not much less than that of actual GSM phones. Thus, the feasibility of implementing a dual-mode terminal such as the one described here is high.

94

Broadband Wireless Mobile: 3G and Beyond

Operator point of view The main items in the operator point of view should be those of the HO (handover) between different cells with different systems, either for UMTS to GSM or vice versa (see Figure 2.37). The handover from a UMTS cell to a GSM cell begins a procedure common to any handover that is the measurement of the parameters of the origin cell and the destination cell. In this case, this measurement of the target cell results in a reservation of resources requirement to the MSC and an acknowledge from the BSS, depending in the services being enjoyed in the origin cell. A special GSM HO command is then initiated by UTRAN and the control is effectively transferred to the GSM BSS. This case is foreseen in UTRAN procedures and puts no additional burden in GSM procedures and takes the same structure as cdma2000. On the other hand, the HO from GSM to UMTS requires modifications in the GSM standards. Modifications are actually being made over the GSM Standard (04.18 Phase 2 1 ) in order to permit this kind of handover. Basically the procedure is symmetric to the one previously discussed. The modifications needed in GSM standards are mainly due to the lack of the UTRAN HO command. cdma2000-IS95 compatibility We switch from interoperability to compatibility because interoperability is a given. cdma2000 was designed as a linear extension to broadband mobile communications of IS95. Either in its MC (Multi Carrier) or DS (Direct Sequence) version, cdma2000 is a upgrade (plus some added features) of the IS95 services.

Figure 2.37

Signalling processes for GSM UMTS interoperability.

UMTS Air Interface

95

User point of view Unlike the UMTS-GSM case, the dual terminal in cdma2000 would reuse much of its hardware (if not all) for IS95 to cdma2000. In this aspect we can expect more compact terminals, less complex and, due to the higher spectral efficiency of the modulation used in cdma2000, higher savings in power terms. Operator point of view A number of advantages make it easy for IS-95 operators to migrate to cdma2000: † Handoffs between systems (in the same frequency, inter-frequency, at cell boundaries or in the same cell) are direct and straightforward. † Possible reuse of system infrastructure (BS, MSC and network) and mechanical infrastructure (buildings, towers, location, rights, etc.). † Possibility of coexistence of TIA/EIA-95-B systems with cdma2000 in the same frequency band via an overlay configuration. † Reuse of billing systems already existing.

All these features and more are the direct consequence that the design of cdma2000 standard relies heavily in backwards compatibility with TIA/EIA-95-B systems.

2.5 Enhancing 3G Capabilities If optimization of the available resources was one of the leading factors in developing the second generation of mobile communications (GSM, IS-54, etc.), in the third generation optimization is a necessity. In mobile communications, spectrum is the most valuable resource, so spectral efficiency is the most critical parameter to optimize. In this context, by spectral efficiency we mean the ratio between the product bit per second served to each user times the number of users served over the bandwidth of the spectrum required to provide such a service. The type of services offered by second-generation systems were in the tens of kbit/s with limited qualities of services. Moreover, the number of users that made up the business plans for second-generation mobile operators were rather limited. However, as has been repeatedly said in this book, the throughput provided in 3G services and the number of users expected in the business plans will be larger than in 2G by far. Under these circumstances, spectrum efficiency will be the key factor in 3G, with much more impact on the operator’s competitiveness than in 2G. Researchers and equipment manufacturers have not been ‘without this reality’ and an amount of work has been done to improve the capacity above the standard achievable with the matched filter, or ‘one-shot optimal’ receiver. The main benefits of using such enhancing techniques are: † Improving the QoS for a given number of users served by the network. When considering real-time/transparent services (including voice), the concept of QoS encompasses data rate and Block Error Rate (BLER). When considering non-transparent data service, the QoS encompass throughput, average delay and blockage probability (quality). † Increasing the maximum number of users served by the network for a given QoS (capacity). † Enlarging the cell range for a given number of users served by a cell and QoS (coverage).

96

Broadband Wireless Mobile: 3G and Beyond

Techniques to improve the spectrum efficiency may be grouped into the following four categories: † Adaptive Antennas (AA): systems using multiple antennas to receive (or transmit) the same information. Instead of using only one antenna to receive (or transmit) the radio signal, combinations of multiple antennas outputs (or inputs) are used to focus the energy towards one direction. The adequate combination of the signals received by the different antennas results in a better link quality for a given transmitted power, or a lower transmitted power required for a given link quality. The same concept of adaptive antennas can be found under different names such as smart antennas, or adaptive beamforming. † Space-Time Transmission Diversity (STTD): systems transmitting the same information coded with different schemes through different antennas. The information is redundantly coded with several schemes and each coded output transmitted through a different antenna. The main difference, sometimes subtle, between STTD and AAs is that the later transmit (or receive) the same signal, except by a complex scale factor, through all the antennas, while the former transmit different signal in different antennas. The complex scale factors applied to each antenna in adaptive antennas are computed in such a way that transmitted energy is focused into an angular sector. In STTD systems, the information data stream is coded into several different signals, and each antenna transmits one of these signals. Because the signal radiated by each antenna element is uncorrelated (orthogonal) to each other, no focalization of the energy is performed, and users in the cell can receive the signal independently of its location, as it happens with an omnidirectional antenna. † Multi-user Detection (MUD): receivers implementing the optimum filter for CDMA scenario. Matched filter theory was proposed in the 1960s as the optimum receiver architecture, however this theory assumes that the only disturbance in the reception is AWGN (Additive White Gaussian Noise). This might not be the situation in some cases, but it is not certain in CDMA systems. A CDMA network capacity is limited by multi-user interference. When the number of users served by a cell increases, the harm caused by the other users’ interference exceed by far the AWGN disturbance, so a matched filter is no longer the optimum receiver. In this scenario, taking into account the statistics of the other users interference, receivers that minimize the overall disturbance power are optimum. Such receivers are known under the generic name of multi-user detection. † Turbo coding: coding schemes that achieve capacities close to the Shannon upper bound. This coding technique concatenates several convolutional codes and interleavers in transmission. An iterative decoding scheme where soft decisions are handed between decoding blocks is implemented in the receiver.

The techniques in these four categories are not exclusive, i.e. several enhancing techniques can be concatenated to add up in performance. For instance, AA can be combined with turbo coding to achieve better performance. There is not a clear border between some of the enhancing techniques groups proposed above, and information on novel enhancing techniques which may create confusion sometimes. This is the case between adaptive antennas and space-time transmission diversity, where some authors propose algorithms of beamforming in transmission and claim this to be a STTD technique. In this case, a simple and clear criterion to decide which group a technique belongs to can be stated. If different antenna elements transmit the same signal except by a complex (amplitude and phase) factor we are dealing with an AA technique,

97

UMTS Air Interface

because the overall effect is to focus the radiated energy towards a particular direction. However, if the correlation between the signals transmitted through different antenna elements is low (or zero), although from coding the same information, thus it is a STTD technique. Similar demarcation problems may occur with the concatenation of AA and turbo coding techniques, what may be confused with STTD. Figure 2.38 summarizes the improvements achieved by each of the enhancing techniques. Of course in this section we do not address enhancing techniques that are inherent in 3G. For example the CDMA multiplexing technique, that it is well known to improve capacity over TDMA under certain circumstances, is not considered as an ‘enhancing technique’ in the 3G context, at least in this book. Neither power control, nor discontinuous transmission for voice links are considered ‘enhancing techniques’. Frequency hopping is not considered either. Although frequency hopping was successfully implemented in phase two of GSM, providing a noticeable capacity increase with a low cost increment, it would not help in a wideband modulation system as proposed for 3G.

Figure 2.38

Summary of enhancing techniques performance improvement.

Enhancing techniques may seem too complex and unfeasible for practical implementation. A careful and critical reading of the technical literature on this issue is required. Although it is true that many papers propose algorithms whose assumptions or complexity make them unfeasible, there are many studies that are perfectly feasible with the present state-of-theart technology. As an example, some proposed ‘blind’ (no pilot/training sequence required) AA algorithms imply a computational load that is not achievable in real-time with present technology. But at the other extreme, some kind of multi-antenna processing (not to say it is an actual AA technique) is already implemented in all the 2G mobile systems: switching among receiving antennas to provide spatial diversity. There is no doubt of the feasibility of simple AA techniques such as ‘beam-switching’; this can be understood as a natural technological evolution from 2G to spatial diversity. Similar argument applies to STTD proposed algorithms. STTD techniques considering convolutional codes and multiple antennas in the

98

Broadband Wireless Mobile: 3G and Beyond

UE should be disregarded as practical for 3G. However space-time transmission diversity techniques using short block codes and two antennas in the Node-B but only one antenna in the UE can be easily implemented. Moreover, they are considered in the standardization process achieved by GPP. Up to this point, the possible advantages of using enhancing techniques have been highlighted. However, there are some disadvantages. The first, and obvious, drawback is the system complexity increase. A second drawback is the need to complicate the standards in order to allow compatibility between equipment that incorporates some mitigation techniques with equipment that does not. The system complexity increase for the four groups of techniques considered above is limited and it has been foreseen as affordable with the present state of the art, at least in some particular environments. Although it is addressed in detail in the following sub-sections, adaptive antennas, multi-user detection and space-time transmission diversity techniques are applied directly to the physical channels, and they do not have any impact on upper layers such as transport channels or logical channels. No additional signalling is required in transport channels or logical channels nor in upper layer protocols. Some extra signalling (pilot signals) are required in the physical layer, in order to be able to estimate the channel. Such pilot signals are transmitted in the Common Pilot Channel (CPICH). Feedback information is sent, if required, from UE to Node-B through the Physical Dedicated Control Channel (PDCCH). Therefore, the inclusion of AA, MUD, or STTD enhancing techniques increases system complexity only in the physical layer of the Uu interface (air interface). The rest of the network will remain unaltered. On the other hand, a turbo coding technique is applied over the transport channels. The information bits in some transport channels may be coded in the transmitter prior to mapping to a physical channel and turbo decoded in the receiver right after being mapped back to a transport channel from the physical channel. The adoption of turbo coding does not require any signalling besides the simple handshaking between ends to agree on its use. In this case, the increase of complexity depends exclusively on the coding and decoding of the transport channels, in the Uu interface. It has no impact at all on the physical layer or on MAC or upper layers. Summarizing for those readers who are not interested on the detail of the different techniques, Tables 2.21 and 2.22 show the complete list of the UMTS FDD transport and physical channels, downlink and uplink, respectively, and the applicability of the different enhancing techniques to each of them. Similar lists can be easily extrapolated for other 3G standards. Finally, before describing in depth the different techniques, note that although the complexity increase expected in the hardware incorporating the enhancing techniques mentioned herein it is not too great, and it is worthwhile noting that such techniques will have an impact on the network planning. These techniques ease problems related to inter-cell interference, intra-cell interference (capacity), fast-fading margin or cell range. Thus, the strategy and tools for the network planning should incorporate changes in order to accommodate the advantages of these techniques. 2.5.1 Adaptive Antennas The basic concept behind adaptive antennas is to change the standard antenna with an

Physical channel

PRACH

PCPCH

DPCCH DPDCH

RACH

CPCH

DCH

Theoretically it could. But it would increase complexity noticeably to achieve little gain. It may be use to serve hot-spots, steering energy towards a group of users. Yes Yes

No

Suitable for AA (reception Node-B)

Applicability of enhancing techniques to UMTS FDD uplink

Transport channel

Table 2.21

No – not applicable because UE is equipped with only one antenna

Suitable for STTD

Yes

Yes

Yes

No

Yes

Yes Yes

Suitable for turbo coding

Suitable for MUD

UMTS Air Interface 99

This channel includes the pilots required for UE to estimate the channel and feed it back to Node-B No – synchronization should be transmitted omnidirectionally for all users within sector No – all these channels are multiplexed, and they should be received by all users within sector

CPICH

CD/CA-ICH CSICH AICH

AP-AICH

SCH

Suitable for AA (transmission Node-B)

Physical channel

Yes

This channel includes the pilots required for UE to demodulate de received signal Yes – but only in the switched mode TSTD

Suitable for STTD

Applicability of enhancing techniques to UMTS FDD downlink

Transport channel

Table 2.22

Yes

Yes

Yes

Suitable for MUD

No

Suitable for turbo coding

100 Broadband Wireless Mobile: 3G and Beyond

P-CCPCH

PICH S-CCPCH

PDSCH

DPCH

BCH

PCH

FACH DSCH

DCH

Theoretically it could. But it would increase complexity noticeably to achieve little gain. It may be use to serve hot spots, steering energy towards a group of users Yes

No – information to be broadcast omnidirectionally for all users within sector No No – no feedback information available from users

Yes

Yes

Yes Yes

Yes

Yes

Yes

Yes Yes

Yes

Yes

Yes Yes

No No

No

UMTS Air Interface 101

102

Broadband Wireless Mobile: 3G and Beyond

omnidirectional radiation pattern for a new one with a directive radiation pattern that is continuously adapted to the environment. The basic idea behind AA is nothing new; engineers in 2G were aware of the capacity gain of using directive antennas and they mostly implemented three antennas covering a 1208 sector each instead of an omnidirectional unique one. The utilization of three sectors multiplied the cell capacity by a factor close to three. (In this case three is an upper bound for the capacity improvement. In a general case with N sectors, the upper bound for the capacity gain is N. Actual figures are always lower than that, but for the case of three sectors of 1208 and uniformly distributed users the actual capacity gain is quite close to the upper bound.) Adaptive antennas go a step further; they provide as many radiation patterns (sectors) as users in the cell. Each radiation pattern, usually named as beam or beamforming, is steered toward one user to maximize its Signal to Interference plus Noise Ratio (SINR). AAs do not maintain fixed beams, but they are adapted continuously to match the dynamics of the environment. Because of its capability of following the users and maximizing their QoS, AAs are also referred to as smart antennas. Because of the duality of the antennas, which states that the transmission and reception behaviour of an antenna are equal, adaptive antennas can be used in transmission or reception. Concept of adaptive antennas In order to implement an antenna with multiple beams and with the capability of dynamically changing the radiation pattern of each beam, AAs are constructed as a digital combination (in base band) of an array of elementary antennas. Because of this implementation, AAs are also named adaptive arrays and the elementary antennas composing the array are called antenna elements. And techniques to compute the optimal combination of the antenna elements are referred to as array processing algorithms. Basic block diagram for AA architecture is depicted in Figure 2.39. It shows the base band processing; if all the radio-frequency and intermediate-frequency processing is obviated, they should appear between the antenna element and the circles that represent a complex value multiplication. In this scheme, let xi(t) denote the signal received at time t by the antenna element i. The AA algorithm defines the weight vector, wi, which determines the radiation pattern of the overall antenna, so the AA output at time t is y(t) (Figure 2.40). If we note the vector x(t) (of dimension N £ 1, where N is the number of antennas) as the vertical stack of the signals received by the N antenna elements 2 xðtÞ

x1 ðtÞ

3

7 6 6 x ðtÞ 7 6 2 7 7 6 6 . 7 6 . 7 6 . 7 5 4 xN ðtÞ

and w a similar stack of the weight vectors

ð20Þ

103

UMTS Air Interface

Figure 2.39

Figure 2.40

Illustration of the adaptive antenna concept.

Block diagram for the implementation of AAs.

104

Broadband Wireless Mobile: 3G and Beyond

2 wðtÞ

w1

3

7 6 6w 7 6 27 7 6 6 . 7 6 . 7 6 . 7 5 4 wN

ð21Þ

then the AA output is computed as y ðt Þ

wH ·xðtÞ

ð22Þ

H

(·) denotes the Hermitian, i.e. vector transposed and conjugated. Now, in order to be able to propose a simple model the following assumptions are made: † The maximum separation between antenna elements is smaller than the product of the light speed times the inverse of the signal bandwidth. Expressed in simpler words, this assumption means the signal envelop suffers a negligible relative delay between any pair of antenna elements, and the only two differences between the signal received by any two antenna elements are an amplitude scale factor and a carrier phase shift. † There is no mutual coupling between antenna elements, i.e. the signal received (or transmitted) by an element does not affect the signal received (or transmitted) by the neighbouring elements. † The number of impinging signals is finite. The three assumptions above are quite realistic for the physical channels and AA proposed for 3G mobile systems. Under these assumptions, the signal vector of the received signals can be modelled as xðtÞ

M X

sj ðTÞ·aj 1 nðtÞ

ð23Þ

j¼1

where x(t) and n(t) are the vector of received signals, also known as ‘snapshot’, and the vector of noise at each element, respectively. sj(t) is the signal transmitted by the j user, and aj is the steering vector of dimension N £ 1 the signal of the j user is received with. aj entries are complex scale factors (amplitude and phase) that are functions of the angle of arrival of the j user. In a general case, the angle of arrival is bi-dimensional: elevation and azimuth. 1,…,N, the expression for the An AA with N elements located on coordinates xj, j steering vector aj is   h      i g1j exp j·kj ·x1 1 w1j ; g2j exp j·kj ·x2 1 w2j ; ::::; gNj exp j·kj ·xN 1 wNj aj a kj kj xn

i 2p 2p h ·^rj · sinuj ·cosfj ; sinuj ·sinfj ; cosuj l l  T xn ; yn ; zn

105

UMTS Air Interface

aj

h

     i g1j exp q1j ; g2j exp q2j ; ::::; gNj exp qNj

ð24Þ

where gij is the gain of the antenna element i towards the direction of the user j. u j and f j are the elevation and azimuth of the j user, respectively. The above notation models a great variety of effects: calibration errors on the antenna element locations, non-omnidirectional antennas (gij), user movement, among others. Without lost of generality, let the first signal s1(t), the signal to demodulate and the rest of the M 1 (total of M users) be the interferences. Desired user is seen by the AA with an angle (bidimensional: elevation and azimuth) q d, then the signals vector received by the AA may be expressed as
 xðtÞ a qd ·d ðtÞ 1 Ai ðqÞ·si ðtÞ 1 nðtÞxðtÞ sðtÞ 1 iðtÞ 1 nðtÞ ð25Þ 1), in which columns are the M 1 where Ai ðqÞ is the matrix of dimensions N £ (M steering vectors of the interferences. Vector si(t), of dimensions (M 1), is the vertical stack of the (M 1) interferences. Under this model, it is well known that the optimum beamforming, w, that maximizes the SINR, and therefore the BER, is
 w Rxx1 a qd ð26Þ where Rxx is the covariance matrix, and (·) h i Rxx E xxH

1

denotes the inverse matrix ð27Þ

The antenna array signals model does not consider multipath effects. However, it is well known that the multipath is one of the limiting factors on the mobile radio link, because of the fading and temporal spreading it causes. In order to incorporate multipath on the signal model, the Vector Channel Impulse Response (VCIR) is defined as hðt; tÞ

LX1



a qi ; fi ·ai ðtÞ·d t

ti



ð28Þ

i¼0


 where qi ; fi ; ai y ti are the elevation, azimuth, complex amplitude and phase of the i path of the multipath channel. The total number of paths in the channel is L. Complex amplitude of a path is defined as

ai ð t Þ

ri ·expðj·ð2p·fi ·t 1 Ci ÞÞ

ð29Þ

where ri is the attenuation suffer by the i path and fi is the Doppler frequency shift of the i path as a result of the possible relative movement in the scenario. Including the VCIR model in the received vector signal in (28), and assuming the stationarity of the channel, x results xðtÞ

LX1



a fi ·ai ðtÞ·s t

 ti 1 iðtÞ 1 nðtÞ

ð30Þ

i¼0

In those situations where the delay spread caused by the channel is much lower than the

106

Broadband Wireless Mobile: 3G and Beyond

inverse of the transmitted signal bandwidth, the above model simplifies into xðtÞ


st

 LX1
 to · a fi ·ai 1 nðtÞ


st

 to ·b 1 iðtÞ 1 nðtÞ

ð31Þ

i¼0

where b is known as the spatial signature b

LX1


 a fi ·ai

ð32Þ

i¼0

Under this model, the optimum beamformer follows the expression in (26), but changing a by b: w


Rxx1 b


ð33Þ

Although up to this point the optimal solution for an adaptive antenna trying to optimize the quality of a given user has been presented, practical implementation cannot be achieved. There are several reasons: † the channel response is not known in the best of the cases, if enough computational power is available, it can be estimated; † estimation of the covariance matrix, Rxx, if computed by the particular implementation, does not match exactly its actual value; † computing the inverse of a N £ N matrix is computationally too expensive, and most algorithms do not use the equation in (25) but they use recursive approaches that asymptotically provide the same solution. Since an optimal solution is not achievable in practical implementations, we will show the particularization that should be made for 3G mobile networks and summarize AA feasible architectures for such scenarios in a later section. Particularities of AA when applied to mobile systems Adaptive antennas implies multiple antenna elements separated enough to receive the same source with different complex scale factors as stated in the models of the previous section. The separation between two consecutive antenna elements should not be too large in order to avoid ambiguity on angles of arrival. Thus, the usual distance between antennas is l /2, being l the carrier wavelength. With these restrictions on mind, and considering the size of the mobile terminals as a fundamental factor on their marketing, the usage of multiple antennas on the UE are disregarded. Therefore adaptive antennas are proposed only for Node-B. However, AA may be used in both links of the radio interface, uplink and downlink. Implementation of AA for the uplink implies the usage of antenna arrays in the receiver of the Node-B. Implementation of AA for the downlink implies the usage of antenna arrays in the transmitter of the Node-B. Assuming the knowledge of the spatial signature as defined in equation (32) is not realistic at all, because it is not stationary as a result of the UE’s mobility. Moreover, the coherence time of a mobile channel, with the UE speed constraints and carrier frequency in 3G, is in the order of few milliseconds, so any weight vector applied in AA for 3G has to be updated every few milliseconds time. We will see in the next section that optimum weight vector can also be computed from a reference signal a pilot that is received by the AA. Then, AA can be implemented in Node-B by adapting itself with the pilot signal received in the Dedicated

UMTS Air Interface

107

Physical Control Channel (DPCCH). The way of adapting an antenna array from a pilot signal is explained in the following section. In the downlink, the Node-B AA works in transmission, therefore there is no way for it to receive a pilot signal. In addition, spatial signature for the downlink cannot be extrapolated from the received spatial signature of the uplink, especially in FDD systems. Therefore UE has to feed back the spatial signature it receives to Node-B. In the UMTS FDD case, this information is fed back to the Node-B in the FeedBack Information (FBI) field of the DPCCH. UE estimates the spatial signature by demodulating the two different pilots contained in the Common Pilot Channel (CPICH) that are respectively sent by each of the two antennas of the Node-B, as specified by 3GPP. Because of the feedback required for AA to work on transmission, this enhancing techniques is named close-loop diversity. Technological alternatives The first alternative that comes when designing AA is the adaptation criterion. By adaptation criterion we mean that we can assume the knowledge of the link to get the optimum weight vector. There are three alternatives here: availability of a reference signal pilot that is continuously received by the AA; the knowledge of the spatial signature, or its real-time estimation for mobile systems; and blind algorithms that use some property (constant modulus, cycle-stationarity, etc.) of the desired signal to estimate the optimum weight vector. Temporal reference or pilot signal The concept of Time Reference Beamformer (TRB) first appeared in [58]. It is based on the knowledge of a reference signal, d(t). This reference signal is be transmitted in parallel with the dedicated physical channels so the AA adapts the weight vector for the AA output to be as close to d(t) as possible. Under this criterion interference and noise are cancelled in a Minimum Mean Square Error (MMSE) fashion. The main drawback of this approach is the need for such a pilot. And theoretically the inclusion of the pilot reduces the spectral efficiency of the system. However, in practical mobile standards this does not mean any inconvenience for the uplink, because dedicated channels always include such a pilot for equalizing purposes. For example, UMTS FDD sends a dedicated physical control channel that contains a quasi-continuous pilot in parallel (quadrature) with the dedicated physical data channel. However, this is a limitation for the downlink. Because in this case the AA works on transmission no pilot signal is available for the beamforming. The cost function to minimize in this approach is the expected value of squared error signal, e(t). The error signal is computed as the difference between the AA output, y(t), and the reference signal, d(t). Then the cost function to minimize can be expressed as i i min E½jdðtÞ wH ·xðtÞ 2 ð34Þ min E½jdðtÞ yðtÞ 2 It is well known that the above minimizing problem is optimized by the weight vector w

Rxx1 ·px

ð35Þ

where px is the correlation vector between the reference signal and the received signal vector, x,

108

px

Broadband Wireless Mobile: 3G and Beyond

  E dp ðtÞ·xðtÞ

ð36Þ

where (·) * means complex conjugate. Spatial reference or knowledge of the spatial signature The optimum combination of antenna elements proposed above is directly applicable if the spatial signature is known by the AA. In this case Applebaum [59] proved that the optimum weight vector given in equation (33) is the one to maximize the SINR. Applebaum also proposed a practical implementation for that AA optimum. This approach is the most suitable for 3G AA working in transmission for the downlink. Because they work on transmission they do not have a reference signal available. TDD systems may try to use a temporal reference beamforming in reception (uplink) and use the same weight vector for transmission (downlink). In this case they have to guarantee the time separation between uplink and downlink is smaller than the coherence time of the channel (inverse of the maximum channel Doppler shift), which is not always the case. FDD systems are required to use spatial reference algorithms for the downlink. Spatial reference algorithms suffer from the need for the spatial signature knowledge. Some researches have proposed algorithms to translate the uplink spatial signature, which can be estimated in Node-B receivers, to a downlink spatial signature. Because of the different propagations that occur through the different frequencies, no proposed technique has been proved to work properly. A second alternative is for the UE to estimate the spatial signature from a pilot signal and to feed it back to the Node-B to adapt the AA. This alternative is the one adopted by 3GPP. Other AA techniques proposed in the literature, such as sidelobe canceller, generalized sidelobe canceller, or minimum variance distortionless response array, are practical implementations of the spatial reference beamforming philosophy. Blind algorithms A third category of proposed AA techniques is known as ‘blind’. Under this category we mean all the AA algorithms that do not require a reference signal or the knowledge of the spatial signature. The criterion they use to adapt the array is to try to recover a known property of the signal at the output of the array. Main exponents of this group of techniques are: constant modulus algorithm, and SCORE family of algorithms. Constant Modulus Algorithm (CMA) adapts iteratively the weight vector to achieve a constant envelope of the AA output signal. Of course this algorithm was developed for constant envelope modulations (BPSK, QPSK, FSK, etc.). SCORE (Self COherence REstoral) family of algorithms work in a similar fashion to CMA but try to achieve some cycle-stationarity properties at the AA output signal. Since all digital modulations are cycle stationary at some specific frequencies, SCORE is, in principle, applicable to any digital system. The main problem of blind algorithms is they have a slow convergence to the optimum weight vector. More so, they may be captured in non-optimal solutions. Because of this slow convergence they cannot be applied in 3G mobile systems where the channel is modelled by fast dynamics. Behind the criterion used by AA to adapt the weight vector, some words should be said on the way to compute the optimums shown above. Basically there are two approaches: closed

UMTS Air Interface

109

form and iterative solutions. By ‘closed form’ we mean the algorithm estimates the different quantities required (Rxx, Rp, b, etc.), and computes the weight vector as proposed in equations (33) or (35). By an iterative approach we mean the algorithm re-computes continuously the vector from the last value of it and the error sequence. Well known examples of the iterative techniques are LMS, BLMS, or RLS. There is nothing established on what kind of technique should be used in 3G AAs. Nevertheless, let us suggest that closed form algorithms are more promising, because they have better performances but higher computational costs. However, some of the quantities to estimate are also needed for channel equalization, so they might be available for the AA. Low complexity approaches Because of the computational cost implied in the solution of equations (33) or (35) and the requirement of re-computing the solution every few milliseconds to be able to follow the channel dynamics, no fully adaptive antenna implementations have been done up to now in commercial applications. (There are pieces of information on AA implementation for military applications but, because of the nature of the information, no great detail on such systems is publicly available.) However, a simplification of the AA concept, midway between antenna diversity implemented in 2G and fully AA, is technologically available at a cost appropriate for its implementation on 3G mobile networks. This simplification is known as Switching Beams Systems (SBS). The conceptual idea of SBS is to have finite set of weight vectors as solution candidates for equations (33) or (35). Then, quality of service achieved by all the candidates is somehow estimated in real-time and consequently the one with the best performance is selected. The computational cost of this approach is proportional to the number of solution candidates in the set of weight vectors. And performance gain is also proportional to the number of possible weight vectors, assuming some minimum requirements on the number of elements and their separation are satisfied by the antenna array. Thus, the great advantage of SBS is its scalability, in the sense it is possible to choose any degree of complexity (cost) increase with its consequent (and proportional) performance increase. Combination of AA and equalizer (RAKE) for CDMA RAKE receivers are proved to improve the Eb/N0 (energy per bit over noise spectral power density) in a multipath environment. Actually, RAKE receivers are the standard way to equalize channel disturbances in CDMA links that combines all the paths in a multipath channel and provides the best Bit Error Rate (BER). The RAKE receiver correlates the input signal with replicas of the same PN code with different relative delays. The correlation done with each replica of the code is known in the literature as RAKE’s finger. Each finger ‘captures’ the energy of one path, assuming a path arrives with the same relative delay of that replica. RAKE receivers improve the overall link quality by optimally combining the output of each finger. Thus, a RAKE receiver can be seen as the optimal temporal combination (filter) to maximize the SNR (Eb/N0 correctly speaking). The concept behind 2D-RAKE receivers is to apply an adaptive antenna to each finger of an array of antennas. Figure 2.41 shows a block diagram for this concept. Optimal spatial combination is performed at each finger prior to combining the fingers. In this way an optimal spatial and temporal combination is done to maximize the SINR, instead of SNR. Although theoretically sound, 2D-Rake receivers imply sophisticated algorithms to

110

Broadband Wireless Mobile: 3G and Beyond

Figure 2.41

Conceptual architecture of 2D RAKE receiver.

compute the finger’s weight vector, which may preclude them from actual implementation in the first stage of 3G network deployment. Practical implementation for 3G Some particularities of AA implementation in different 3G standards are now addressed. UMTS FDD As mentioned in a previous section, implementation of AA in the UE is not considered for practical and obvious reasons. Thus AA implementation is only considered at the Node-B end. Therefore AA will work on reception for the uplink and in transmission for the downlink. Uplink AA will estimated the optimum weight vector (or set of them if implementing 2DRake receivers) based on temporal reference beamformings. In this case no cooperation between UE and Node-B is needed, and therefore no specifications are expected from any standardization institute. Downlink AAs are a different case. They work on transmission; therefore they cannot use reference signals, so they will operate using spatial reference algorithms. In this case, cooperation between the UE and Node-B is required.

UMTS Air Interface

111

3GPP September 2000 release, has standardized the following issues: implementation of AA is not mandatory; in any case, UE shall have the capacity of measuring the spatial signature from two transmitting antennas in Node-B. This measurement will be done over the two different pilot signals transmitted by the two antennas (case of Node-B implementing AA) contained in the Common Pilot Channel (CPICH). This information will be fed back to Node-B in the FBI field of the DPCCH. With this standard on mind, the following conclusion can be stated: † when implementing AA in transmission, they will have two antenna elements; † in the downlink AAs will be implemented (basically) for dedicated channels; † AAs working on reception (uplink) may incorporate any number of antenna elements.

The more elements AAs provide the more capacity gain, and the more complexity the greater the cost. For symmetry with the downlink, two antenna elements may be expected. Downlink As mentioned before, AAs will impact only on physical channels, and there will not be signalling associated apart from the mentioned pilots in the CPICH and the FBI field in the DPCCH. Figure 2.42 shows the channels that AAs will impact on.

Figure 2.42

Impact of adaptive antennas on the UMTS FDD downlink channel structure.

The transmitter structure to support transmission AA, for DPCH, is shown in Figure 2.43. Channel coding, interleaving and spreading are done as in the non-diversity mode. The spread complex-valued signal is fed to both TX antenna branches, and weighted with antennaspecific weight factors w1 and w2. The weight factors are complex-valued signals (i.e. wi ai 1 jbi), in general. These weight factors are calculated on a per slot and per user basis. The weight factors are determined by the UTRAN. Uplink Implementation of AAs in the uplink will not involve any signalling to pass any layer above the physical channels. Figure 2.44 shows the channels which AAs will impact on.

112

Broadband Wireless Mobile: 3G and Beyond

Figure 2.43 Downlink transmitter structure to support transmit diversity for DPCH transmission (UTRAN access point). 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

Figure 2.44

Impact of adaptive antennas on the UMTS FDD uplink channel structure.

UMTS TDD The only difference from applying AA to FDD to TDD is the possibility of extrapolating uplink beamforming to the downlink. The appropriateness of this extrapolation depends on the slot assignment. Contiguous slot assignment for uplink and downlink would allow the extrapolation, but slots separated by five or seven slots would lead to bad performances. Although the multiplexing technique in TDD is different from the one in FDD, the channel structure is quite similar. Some channels are added to provide Opportunity Driven Multiple

UMTS Air Interface

113

Access (ODMA), but the majority of the channels remain the same. Therefore all conclusion on AAs applied to FDD can be extrapolated to the TDD mode. Performance analysis Network performances are enhanced by the use of AAs in the following parameters: † Larger cell coverage: an AA of M elements exhibits an antenna gain that can be approximated by M times larger than the antenna elements gain. This approximation is particularly good when working close to the ‘broadside’ region of the antenna, which is usually the case in sectorized antennas. An antenna gain increase by a factor M means a cell range increase by a factor M1/g , where g is the exponent of the propagation loss law (typically modelled by g 4). So, the required number of Node-Bs to cover a given region decreases by a factor of M2/g . † Capacity increase for a given infrastructure: understanding the capacity as the density of users per area unit (users/km 2). The usage of AAs allows the concept of Reuse Within the Cell (RWC) or Space Division Multiple Access (SDMA) which basically means the same resources may be used within the same cell. Assuming a user-uniform angular distribution of the users along the antenna sector, the usage of an AA of M elements enlarges the capacity by a factor M. CDMA systems exhibit a capacity that is interference limited. The usage of an AA of M elements enlarges the capacity by a factor M too, of course, if a uniform angular distribution of users can be assumed. † Quality of service increase for a given infrastructure and capacity: if range and number of users are kept constant, the usage of AA improves the SINR, and consequently decreases the Block Error Rate (BLER), which impacts directly on throughput and average delay.

2.5.2 Space-time transmission diversity The concept of Space-Time Transmission Diversity (STTD) comes from the extrapolation of Receivers Antenna Diversity widely used in 2G. The idea of receiver antenna diversity is to place two (several in general) antennas in different locations and select the output of the one with the best SINR. Placing two antennas in reception defines two different radio channels, where a channel is the radio-wave propagation between transmitter antenna and each of the receiving antennas. If the two receiving antennas are separated enough, and for this purpose 10 times the wavelength is enough, the statistics of the two channels are independent. Therefore, the probability of having a deep fading on both channels simultaneously is the square of the probability of having a deep fading on one of them. In consequence the link quality is roughly speaking improved by 3 dB. Now assuming two transmitting antennas are placed at relative distant locations, when trying to extrapolate the concept of antenna diversity to the transmitter two alternatives can be followed: † Antennas transmit alternatively and receiver checks the quality each transmitter antenna is received with. This information is somehow fed back to the transmitter who decides consequently which antennas transmits through. This approach is known as Time Switch Transmit Diversity (TSTD); UMTS practical implementation of this approach does not

114

Broadband Wireless Mobile: 3G and Beyond

use a feedback channel, but it extrapolates the levels of received signal by the two antennas to transmission. In other words, when UMTS uses TSTD, it transmits through the antenna with higher level of received signal. † Both antennas transmit simultaneously the same information but coded by different coding schemes. This approach is known as Space-Time Transmit Diversity (STTD). In this approach two different coding schemes should be necessarily used for each antenna. If the same signal would be transmitted simultaneously through the two antennas, the overall system would behave as a phased array with a zero degrees relative phase between antennas. The two antennas, viewed as a phased array, would produce a narrower antenna beam, and it would leave some UEs out of coverage. In this approach there is no need for feedback. This is the reason why 3GPP named STTD as open loop diversity. 3G mobile systems may implement STTD techniques. The September 2000 release of 3GPP states that implementation of STTD will be optional for Node-B, but UE has the functionalities to support it. In the case of implementing transmission diversity, most physical channels will implement the STTD approach, except those channels that have to be demodulated prior to getting any knowledge from Node-B, and therefore before the UE can be informed about whether that cell is using STTD or not. Those channels that cannot implement STTD are Common Pilot Channel (CPICH) and Synchronization Channel (SCH). Implementation of TSTD is an option for these channels. Figure 2.45 shows a generic transmitter for a channel implementing STTD-enhancing techniques. Figure 2.46 shows the scheme of the receiver. The basis for the performance improvement achieved by STTD is that two (different) signals containing the same information are simultaneously, through channels h0 and h1, statistically independent. The receiver combines properly the two received signals to recover the original information. Because the probability of having a deep fading simultaneously is much lower than the probability of having a deep fading in only one channel (the probability of simultaneous fading is the product of probabilities of having fading in each channel), the recovered information fading has better performance statistics. The above rationale pretends only to provide an intuitive explanation; a later section will show the mathematical proof of this quality improvement.

Figure 2.45 Generic scheme for a STTD transmitter.

115

UMTS Air Interface

Figure 2.46

Generic scheme for a STTD receiver.

A great deal of research has been done in this field. Many results available in the literature address the problem of having multiple antennas in transmission and multiple antennas in reception, the systems known under the generic name of Multiple Input Multiple Output (MIMO). These kind of systems will not be addressed in this section since they are not practically implemented in 3G mobile networks because of the multiple antennas required in the UE. Of course, results on MIMO research are more promising, improvement-wise, than STTD having multiple antennas on transmission only. Moreover, some of the studies found in the technical literature analyse the MIMO systems as the combination of a STTD system in the transmitter plus an AA system in the receiver, concluding that the MIMO gain is the addition of STTD gain plus the AA gain. A brief historical reference The first efficient STTD scheme was proposed by Witteneben [54], which included the transmit diversity scheme proposed by Seshadri and Winters [55] as a particular case. Seshadri and Winters’ proposed system reproduces the scheme in Figure 2.45 but replaces the channel coder and the serial to parallel converter by a block with two outputs, one connected directly to the input and the second being a delayed version of the input. Later Foschini introduced a multi-layer architecture for STTD [56]. In this context, ‘efficient scheme’ means a ratio of bit rate and bandwidth (bps/Hz) close to one. Recently space-time coding schemes for multiple transmitting antennas and receiver with a moderate signal processing capacity have been proposed [50]. These new coding schemes show a larger gain than schemes in [54] and [55]. Proposed convolutional (trellis) codes have

116

Broadband Wireless Mobile: 3G and Beyond

quite a good performance in slow fading environments, and achieve up to 3 dBs of gain when using two transmitting antennas, as analysed by Foschini y Gans [57]. In order to reduce receiver complexity, Alamouti describes a simpler scheme for spacetime block coding using only two transmitting antennas [51]. Alamouti’s codes were generalized for an arbitrary number of antennas in [52] and [53]. The codes in these last references have the property of requiring simple maximum likelihood receivers, while they provide the maximum achievable gain when applied to one-dimension constellations [50]. In [52] the space-time block codes with simple implementation and good performances are proposed for two-dimension (complex) constellations. Technological alternatives There are several alternatives for implementing transmit diversity. For example, some arguments may be given on the alternative between TSTD (switching transmission) and STTD (simultaneous transmission). The differences between them have been outlined in a previous section. In this section we concentrate in the two main options available for implementing STTD systems: block codes and convolutional (trellis) codes. Block codes Let us use the Alamouti [51] scheme as a starting point. Alamouti’s STTD scheme involves three processes: coding at the transmitter, combining at the receiver, and a maximum likelihood detector after the combiner. Each of the three stages is now addressed in detail. 1. Coding at the transmitter. Every symbol period, T, each transmitting antenna sends its own symbol. Let s0 and s1 be two consecutive information symbols to transmit, in Alamouti’s system, during time interval [0 T] antenna 0 transmits symbol s0, and antenna 1 transmits s1*, and antenna 1 transmits symbol s1. During interval [T 2T] antenna 0 transmits symbol * symbol s0 . Table 2.23 shows this coding scheme. Table 2.23 Coding scheme in Alamouti’s STTD transmitter where (·) * means the complex conjugate

Time [0 T] Time [T 2T]

Antenna 0

Antenna 1

S0

S1 S*0

S*1

Assuming the two channels, h0 and h1 as defined in Figure 2.46, are stationary during the time interval [0 2T], the signals at the receiving antenna are r0

rðtÞjt[½0




r1

rðtÞjt[½T




h0 s0 1 h1 s1 1 n0 h0 sp1 1 h1 sp0 1 n1

ð37Þ ð38Þ

where n0 and n1 are different, and independent, realizations of the Additive White Gaussian Noise (AWGN).

117

UMTS Air Interface

Channel responses, h0 and h1, contain amplitude response, a 0 and a 1, and phase response, f 0 and f 1: h0

a0 ejf0

ð39Þ

h1

a1 ejf1

ð40Þ

2. Combining at the receiver. The receiver in Figure 2.46 estimates the channels, h0 and h1, and combines the signals received during two consecutive symbol periods in the following way: s 00

hp0 r0 1 h1 r1p

ð41Þ

s 01

hp1 r0

ð42Þ

h0 r1p

Substituting equations (37) and (38) into (41) and (42) results in   s 00 a20 1 a21 s0 1 hp0 n0 1 h1 np1 s 01



 a20 1 a21 s1 1 hp1 n0

h0 np1

ð43Þ ð44Þ

3. Maximum likelihood detector. Signals in equations (43) and (44) correspond to a standard modulation where the amount of noise doubled by the statistics of the channel attenuation is much more favourable. A good model for a 0 and a 1 is, in most of the mobile links scenarios, a Rayleigh random variable. In this case, a20 1 a21 is modelled by a x 2 random variable, which presents shallower fading. From equations (43) and (44) and using the well-established theory for the optimum receiver with minimum probability of error, it can be concluded that implementing a maximum likelihood detector after the combiner will provide the best bit error rate. Table 2.23 defines the following coding matrix: ! x1 x2 ð45Þ xp2 xp1 The number of columns is equal to the number of antennas and the number of rows is equal to block code length. In order to have good performance at the same time as simple and linear implementation, matrix columns have to be orthogonal. The goodness of coding scheme in Table 2.23 resides on its spectral efficiency, 1 bps/Hz. This efficiency figure means the STTD scheme does not increase the symbol rate. In [52] a generalized procedure to obtain STTD coding matrixes with dimensions larger than 2 £ 2 is proposed. In [52] it is concluded that for a number of antennas larger than 2 (more than 2 columns) it is not possible to find orthogonal and square matrixes, but the number of rows has to be larger than the number of columns in order to guarantee orthogonality. This type of rectangular matrix leads to space-time coding schemes with output code blocks longer than the input ones, therefore spectral efficiencies lower than 1 bps/Hz. For three and four transmitting antennas it is possible to find schemes with efficiency of 0.75 bps/Hz.

118

Broadband Wireless Mobile: 3G and Beyond

Convolutional codes Similar to STTD systems based on block codes, convolutional (or trellis) coding is based on the approach outlined in Figure 2.45. The usage of convolutional codes for STTD systems provides good performances up to 2 or 3 dB from the maximum theoretical capacity [57]. However their implementation implies a high complexity where a multidimensional Viterbi decoder is required. Also estimation of the channel becomes a more challenging task than it is for block codes. Although models for trellis STTD systems have been proposed for 4-PSK, 8PSK and 16-QAM modulations, and quite good performances have been shown in simulations, the complexity of these schemes will prevent them from practical implementation at least in the first stages of 3G mobile networks. Practical implementation for 3GThere is no strong reason to not implement STTD simple schemes in 3G downlink for the following three reasons: † complexity involved in block coding techniques with code length and number of antennas equal to two is affordably small; † it does not require any change on upper layers remember no signalling feedback is required (open loop diversity); † improvement of average Eb/N0 is close to 3 dB with the consequent capacity and range improvements. UMTS FDD 3GPP September 2000 release defines a STTD block coding scheme with two antennas and a block code of four bits (two QPSK symbols). 3GPP documents state that implementation of STTD will be optional for Node-B, but it will be mandatory for UEs to support this mode. To help UEs in the channel estimations, Common Pilot Channel (CPICH) includes two different pilot signals, one per antenna, where Node-B incorporates STTD.

Figure 2.47

Impact of STTD on the UMTS FDD downlink channel structure.

UMTS Air Interface

119

Figure 2.48 Downlink transmitter structure to support transmit diversity for SCH transmission (UTRAN access point). 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

Downlink Implementation of STTD will not involve any signalling to pass any layer above the physical channels. Figure 2.47 shows the channels which STTD will impact on. Transmit diversity for SCH Time Switched Transmit Diversity (TSTD) can be employed as a transmit diversity scheme for the synchronization channel. The transmitter structure to support transmit diversity for SCH transmission is shown in Figure 2.48. P-SCH and S-SCH are transmitted from antenna 1 and antenna 2 alternatively.

Figure 2.49 Block diagram of the transmitter (STTD). 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

120

Broadband Wireless Mobile: 3G and Beyond

Figure 2.50 Block diagram of Block STTD encoder The symbols Si are QPSK, N is the length of the block to be encoded. 3GPP TSs and TRs are the property of ARIB, CWTS, ETSI, T1, TTA and TTC who jointly own copyright in them. They are subject to further modification and are therefore provided to you ‘as is’ for information purposes only. Further use is strictly prohibited.

Transmit diversity for P-CCPCH Block Space Time Transmit Diversity (Block STTD) may be employed as a transmit diversity scheme for the Primary Common Control Physical Channels (P-CCPCH). The open loop downlink transmit diversity employs a Block Space Time Transmit Diversity scheme (Block STTD). A block diagram of the Block STTD transmitter is shown in Figure 2.49. Before Block STTD encoding, channel coding, rate matching, interleaving and bit-to-symbol mapping are performed as in the non-diversity mode. Block STTD encoding is separately performed for each of the two data fields present in a burst (each data field contains N data symbols). For each data field at the encoder input, two data fields are generated at its output, corresponding to each of the diversity antennas. The Block STTD encoding operation is illustrated in Figure 2.50, where the superscript ( *) stands for complex conjugate. If N is an odd number, the first symbol of the block shall not be STTD encoded and the same symbol will be transmitted with equal power from both antennas. After Block STTD encoding both branches are separately spread and scrambled as in the non-diversity mode. The use of Block STTD encoding will be indicated by higher layers. Uplink Uplink will not implement STTD in any of its forms because the existence of multiple antennas is not contemplated for UE. UMTS TDD There is no practical difference from applying STTD to FDD than to TDD. Operational principles and improvement factors will be the same for both modes. Although the multiplexing technique in TDD is different from the one in FDD, the channel structure is quite similar. Some channels are added to provide Opportunity-Driven Multiple Access (ODMA), but all the conclusions in applying STTD to the FDD mode can be applied to TDD in a straightforward manner. Performance analysis This section addresses the performance improvements achieved when using STTD. The analysis is done for the particular implementation described as an option in the 3GG September 2000 release: two transmitting antennas and only one antenna in the receiver with a code block length of 2 symbols.

121

UMTS Air Interface

Figure 2.51

STTD Gain.

In our analysis, the following model is assumed: rt

2 X

hi cit 1 ht

ð46Þ

i¼1

where cti are the transmitted symbols through the antenna I at the instant t, which are assumed to have unitary energy, and h t are independent realizations of a Gaussian random variable of zero mean and variance:

s2h

1 2SINR

ð47Þ

The coefficients hi are modelled as complex Gaussian independent variables with zero mean and 0.5 variance on real and imaginary parts, respectively. Extensive simulations have been performed to analyse the BER performance increases. Some detailed simulation and analysis can be found for generic STTD block schemes. Figure 2.51 presents results for the UMTS scenario described above. As can be seen, the improvement achieved by STTD depends on the QoS objective. As example, for a BER of 10 3 the required transmitted power is around 10 dB lower when implementing STTD, with the consequent coverage and capacity increase. And note that only two antennas are required, which means doubling the radio frequency hardware, but no signalling is passed and therefore no complexity is added besides the UE demodulator. Figure 2.51 includes a quite simple channel model. A more realistic channel model should be included in the simulation to obtain a more accurate gain estimation. 2.5.3 Turbo coding A new coding scheme proposed in 1993 by Berrou, Glavieux and Thitimajshima at the International Conference on Communication in Geneva [13], claimed that a combination of parallel concatenated convolutional codes and iterative decoding could provide reliable

122

Broadband Wireless Mobile: 3G and Beyond

communications at a signal-to-noise ratio that is within a few tenths of a dB of the theoretical Shannon limit [14]. This performance increase has had a strong effect on wireless personal communications systems, where the bandwidth demands are a consequence of increased data services. A turbo code can be thought of as a refinement of a concatenated encoding structure, first proposed by Forney [15], and an iterative algorithm for the decoding of the associated code sequence. The codes are constructed by applying two or more component codes to different interleaved versions of the same information sequence. The optimum receiver in that situation cannot perform its decisions on a symbol-by-symbol basis, so that decoding on a particular information symbol uk involves processing a whole portion of the received signal. Additionally, in order for a concatenated scheme such as turbo code to work properly there are two main points. First the decoding algorithm should not limit itself to passing ‘hard’ decisions among the decoders; to best exploit the decision learned from each decoder the decoding algorithm should exchange soft decisions rather than hard ones. And second, for a system with two or more components codes, the concept behind turbo decoding is to pass soft decisions from the output of one decoder to the input of the following one, and iterate this process several times to produce better decisions. State of the art Concatenated coding scheme ideas (a class in which you can include product codes, multilevel codes, generalized concatenated codes and serial and parallel concatenated codes) obtain significant coding gains from relatively simple component codes. Concatenated coding schemes were first proposed by Forney [15]. The resulting coding scheme allows a relatively simple decoding scheme based on stage decoding. The turbo coding proposal set in motion several research projects to reproduce the results claimed by the authors and find the theoretical aspects that define the working base of the scheme [16 19]. In the general effort of understanding why turbo codes work as they do, several research teams focused on the soft-in/soft-out decoders that form the building block of the turbo decoders [29]. It was soon noted that the turbo decoding duplicated an algorithm for propagation in graphs [20,21] that lead to a reinterpretation of the decoding problem and greater insights of how the turbo decoder works. In a related area it has been noted that the decoding problem can be interpreted in terms of parameter estimation for hidden Markov models [22]. However, there are still several open topics. Some [16] have in one way or another been treated in other papers recently; others are still under study. Analytical bounding techniques [16,24] have been developed for AWGN channel, but the need for analytical results for bounding techniques applied to fading channels [16] have been pointed out. As a first approach to this requirement, several papers have addressed the study of Rayleigh channels from the simulation point of view [25 28]. Several iterative algorithms have already been proposed, a comparison of them, as well as reducing the number of iterations and consequently the delay, would be an important achievement. At this point in time there are restrictions that apply to real-time implementation of decoding algorithms like the proposal for turbo coding. The algorithm is intrinsically suboptimal from the point of view of the implementation in a real-time working system. First,

123

UMTS Air Interface

Figure 2.52

Turbo decoder block diagram.

there is the delay associated with the frame reception when decoding and with the iterative algorithm that improves the performance. Secondly the computational complexity which, in this case, is increased highly with the number of operations needed to decode one symbol. Also related to the memory requirements, which is another important issue in real time implementation, we should take into account the stage decoding. Stage decoding defines a structure in which each of the constituent decoders process a part of the information. The algorithm in each decoder requires information related to the whole frame when processing one symbol. Additionally the information stored will be truncated resulting in a precision that depends on the implementing device. Depending on the number of bits used to store each sample of information, the memory requirements could be increased. It should be noticed that the performance of the decoding algorithm may be affected by the round-off carried out; thus there is an important trade-off between performance and memory requirements. It is a fact too that the performance increases with the interleaver length and that the interleaver choice is not really critical. But the importance of short interleavers related to delay decoding sensitivity makes the study of the interleaver length important especially under the presence of fading. Some results have already been obtained in this direction in [23]. Decoding algorithm The decoder is made up of two elementary decoders, the interleaver and the de-interleaver in a serial concatenation scheme (Figure 2.52). There exists a correlation between the encoded symbols that extends beyond the symbol period, provided by the coding process. The optimum receiver in that situation cannot perform its decisions on a symbol-by-symbol basis, so that decoding on a particular information symbol uk involves processing a whole portion of the received signal. The decision rule can be either optimum with respect to a sequence of symbols, or with respect to the individual symbol uk. The most widely applied algorithm for minimizing the probability of sequence error for convolutional codes is the Viterbi algorithm [49]. Optimum individual symbol decision algorithms must base their decisions on the maximum a posteriori (MAP) probability. Individual

124

Broadband Wireless Mobile: 3G and Beyond

symbol decision algorithms yield a better performance in terms of symbol error probability than, for example, the Viterbi algorithm, but they present a much higher complexity. Concatenated coding schemes [15] define that the optimum output of the preceding decoder towards the following one should be in the form of the sequence of the probability distributions over the code alphabet conditioned on the received signal, that is the a posteriori probability distribution. Similarly in order for a concatenated scheme such as turbo code to work properly, the decoding algorithm should not limit itself to passing hard decisions among the decoders. To best exploit the decision learned from each decoder the decoding algorithm should exchange soft decisions rather than hard ones. The Logarithm of Likelihood Ratio (LLR) associated with each decoded bit will be the relevant piece of information for the following decoder, since P(uk i j observation) is the a posteriori probability of the data bit uk. Lðuk Þ

log

Pðuk Pðuk

1uobservationÞ 0uobservationÞ

ð48Þ

The additional concept behind turbo decoding is to pass soft decisions from the output of one decoder to the input of the following one, and iterate this process several times to produce better decisions. In the encoder scheme (Figure 2.52), we consider that all the information bits obtained from the encoder are sent through the channel, the subscript kcorresponds to the time sequence, and s and r identify the systematic received bit and the recursive one, respectively. In the decoder scheme (Figure 2.52) we have avoided the time index, since the decoding process is iterative. l1k and l2k are defined as the LLR of the probability transitions of the channel computed by the soft demodulator: ! !3 2 pðys1k uuk 1Þ pðyr1k uxr1k 1Þ ; log " # 6 log 7 6 pðys1k uuk 0Þ pðyr1k uxr1k 0Þ 7 l1k 7 6 ð49Þ lk 6 ! !7 6 l1k pðys2k uxs2k 1Þ pðyr2k uxr2k 1Þ 7 5 4 ; log log pðys2k uxs2k 0Þ pðyr2k uxr2k 0Þ

Figure 2.53

The transmission system.

125

UMTS Air Interface

Figure 2.54

Notation within the trellis.

The whole transmission system is described in Figure 2.53, where # " # " s r # " y1k ðuk 1 ns1k ; xr1k 1 nr1k Þ ðy1k ; y1k Þ yk y2k ðys2k ; yr2k Þ ðxs2k 1 ns2k ; xr2k 1 nr2k Þ

ð50Þ

MAP and modified BCJR algorithm for optimal decoding of RSC codes The BCJR algorithm is the optimum algorithm to produce the sequence of APP. We call this algorithm the BCJR algorithm for the authors’ initials [48]. The BCJR algorithm is going to be used for each of the decoders in Figure 2.52. Each decoder, and consequently each algorithm, will use respectively l1k ðls1k ; lr1k Þ and l2k ðls2k ; lr2k Þ. So when describing the algorithm in this section, and using the notation yk we are going to refer to the pair yk ðys1k ; yr1k Þ or yk ðys2k ; yr2k Þ (either of them) instead of the notation yk described by Figure 2.54 and equation (50). It is going to be necessary to make some modifications within the BCJR algorithm given the recursive nature of the RSC [13]. The BCJR algorithm examines the received sequence of symbols and estimates the APP of the trellis states and transitions of the encoder. In a nonrecursive convolutional code for every state sequence there is a unique and easily determined path through the trellis diagram, so the APP of the input symbol conditioned to the received sequence can be easily obtained. This is not the case for a RSC code. The following notations are going to be used referring to each edge e of the trellis: † the starting state: Ss ðeÞ † the ending state: Se ðeÞ † the input symbol: uðeÞ

The APP of a decoded bit can be derived from the conditional probability lk ðeÞ defined by

lk ðeÞ

Pðuk

uðeÞuyÞ

ð51Þ

126

Broadband Wireless Mobile: 3G and Beyond

where y is the received sequence of symbols ðy
1 ; y2 ; …; yN Þ. X Pðuk iuyÞ lk ðeÞ

ð52Þ

e:uðeÞ¼i

In order to compute the joint probability lk ðeÞ, the following functions are defined:

lk ðeÞ

ak 1 ðSs ðeÞÞgðck ðeÞÞbk ðSe ðeÞÞ

ak ðsÞ

Pðsk

ð53Þ

X

suðy1 ; y2 ; ::; yk ÞÞ

ak 1 ðSs ðeÞÞgðck ðeÞÞ

ð54Þ

e:Se ðeÞ¼s

bk ðsÞ

Pðsk

suðyk11 ; yk12 ; ::; yN ÞÞ

X

bk11 ðSe ðeÞÞgðck11 ðeÞÞ

ð55Þ

e:Ss ðeÞ¼s

gðck ðeÞÞ

pðyk uuk

uðeÞ; sk

1

Ss ðeÞ; sk

Se ðeÞÞ

pðysk uuk

uðeÞÞpðyrk uxrk

xr ðeÞÞ

ð56Þ

The ak ðsÞ and bk ðsÞ values can be recursively calculated, their initial values provided. The first one is calculated with a forward recursion within the trellis, knowing the initial value: ( 1 if s s0 ð57Þ a0 ðsÞ 0 if s – s0 And bk ðsÞ is given by a backward recursion within the trellis, using the initial value ( 1 if s s0 bN ðsÞ 0 if s – s0

ð58Þ

The different steps of the modified BCJR algorithm are: 1. Initialize a0 ðsÞ and bN ðsÞ according to equations (57) and (58). 2. As soon as each term yk of the whole sequence y is received, the soft demodulator provides g (c(e)) and the decoder computes the probabilities ak ðsÞ according to equation (54). All the values of ak ðsÞ and g (c(e))obtained are stored. 3. When the entire sequence y ðy1 ; y2 ; …; yN Þ is received the decoder recursively computes the probabilities bk ðsÞ according to equation (55), and uses them together with the stored values of ak ðsÞ and g (c(e)) to compute lk ðeÞ and finally the APP of the decoded bit. MAP decoding for the PCCC: iterative decoding In the concatenated decoding scheme proposed in Figure 2.52, each of the decoders is going to compute the APP Pðuk uðy11 ; y12 ; :::; y1N Þ; u~1k Þ and Pðuk uðy21 ; y22 ; :::; y2N Þ; u~2k Þ, respectively or equivalently the LLR using the BCJR algorithm: L1 ðuk Þ

log

Pðuk Pðuk

1uðy11 ; y12 ; :::; y1N Þ; u~ 1k Þ 0uðy11 ; y12 ; :::; y1N Þ; u~ 1k Þ

ð59Þ

L2 ðuk Þ

log

Pðuk Pðuk

1uðy21 ; y22 ; :::; y2N Þ; u~ 2k Þ 0uðy21 ; y22 ; :::; y2N Þ; u~ 2k Þ

ð60Þ

The symbols u~ 1k and u~ 2k are the new information provided by each decoder in the decoding

127

UMTS Air Interface

process. They are also called ‘new data estimates’, ‘innovations’ or ‘extrinsic information’ [17]. And as seen in equations (59) and (60), they are going to be used to generate the APP of the information sequence in the iterative process. The following definitions are considered as the first step to demonstrate how each decoder generates the innovations for the next decoder in the iterative process: L~ 1 ðuk Þ

L~ 2 ðuk Þ

log

Pðuk Pðuk

1uu~ 1k Þ 0uu~ 1k Þ

log

Pðuk Pðuk

1uðy21 ; y22 ; :::; y2N Þ; ðu~21 ; u~22 ; :::; u~2k 0uðy21 ; y22 ; :::; y2N Þ; ðu~21 ; u~22 ; :::; u~2k

log

Pðuk Pðuk

1uu~ 2k Þ 0uu~ 2k Þ

log

Pðuk Pðuk

1uðy11 ; y12 ; :::; y1N Þ; ðu~11 ; u~12 ; :::; u~1k 0uðy11 ; y12 ; :::; y1N Þ; ðu~11 ; u~12 ; :::; u~1k

~ 2k11 ; :::; u~ 2N ÞÞ 1; u ~ 2k11 ; :::; u~ 2N ÞÞ 1; u

~ 1k11 ; :::; u~ 1N ÞÞ 1; u ~ 1k11 ; :::; u~ 1N ÞÞ 1; u

ð61Þ

ð62Þ

An important point to consider is that, when decoder 1 generates L~ 2 ðuk Þ for decoder 2 it should not contain any information about u~1k which had been generated by decoder 2 itself. The same way when decoder 2 generates L~ 1 ðuk Þ for decoder 1 it should not contain any information about u~2k which had been generated by decoder 1 itself. This comes from the idea that the fundamental principle for feeding back information to another decoder is never feeding back information to another decoder that stems from itself. This is considered in the definition of both and in the block diagram in Figure 2.52, where we made the following statements that can be demonstrated based on the independence between the extrinsic information and the input sequence for each of the decoders: L~ ðmÞ 1 ðuk Þ

LðmÞ 2 ðuk Þ

L~ ðm 2

1Þ

ðuk Þ

ð63Þ

L~ ðmÞ 2 ðuk Þ

LðmÞ 1 ðuk Þ

L~ ðm 1

1Þ

ðuk Þ

ð64Þ

where the index m supports the iteration steps. The aim of each decoder is then to compute L1(uk) and L2(uk), respectively, and from that information directly extract the information provided by the other decoder in the previous iteration (equations (63) and (64)), and feed the ‘clean’ innovation information to the other decoder. The L1(uk) and L2(uk) are going to be obtained by applying the BCJR algorithm, but instead of using the sequence ððysi1; yri1 Þ; ðysi2; yri2 Þ; …; ðysiN; yriN ÞÞ for the computations, we are going to use ððysi1; yri1 ; u~ i1 Þ; ðysi2; yri2 ; u~ i2 Þ; …; ðysiN; yriN ; u~ iN ÞÞ. Turbo coding schemes for 3G Two different schemes have been proposed for 3GPP and 3GPP2. The first difference is the coding rate; in 3GPP proposal the only possible rate is 1/3 and is achieved by puncturing the concatenated output of two parallel recursive systematic convolutional codes of rate 1/2. 3GPP2 proposes different coding rates 1/2, 1/3 and 1/4. The different coding rate is achieved with different puncturing for a parallel concatenation of two 1/3 recursive systematic convo-

128

Broadband Wireless Mobile: 3G and Beyond

Figure 2.55

Turbo coding scheme for 3GPP.

lutional codes. The resultant coding scheme for puncturing for 1/3 is the same as used for the 3GPP proposal (Figures 2.55 and 2.56).

2.5.4 Multi-user detection With the introduction of spread spectrum communications, the problem of having a reliable detection of the data from multiple users was raised. The once generalized belief that the interference caused by other users in the communication channel was very similar in its characteristics to Gaussian noise has been discarded. Thus, the optimum receiver in a white Gaussian noise channel, i.e. the matched filter, is proved to be highly unreliable in a multi-user environment. This is because this optimum receiver is not the optimum in a multiuser environment and the optimum receiver is one that takes into account the interference caused by the users which is different to the signal of interest. This problem can be illustrated with a simple example. Let us consider a simplified baseband multi-user environment in which the K users are synchronous. The received signal in the antenna is rðtÞ

A1 b1 s1 ðtÞ 1

K X

Ak bk sk ðtÞ 1 nðtÞ

ð65Þ

k¼2

where Ak, bk and sk(t) are the signal amplitude, the bit value and the spreading sequence of user k, respectively. The noise amplitude is represented by nðtÞ where noise is an i.i.d.

129

UMTS Air Interface

Figure 2.56

Turbo coding scheme for 3GPP2.

Gaussian process. The result after unspreading with the sequence of the user of interest (say user 1), is yðtÞ

s1 ðtÞrðtÞ

A1 b1 1

K X

Ak bk r1k ðtÞ 1 s1 ðtÞnðtÞ

ð66Þ

k¼2

where r1k ðtÞ is the correlation of the signature sequence of the user of interest with that of user k. From the formula above it is clear that a potentially catastrophic situation can develop. If A1 b 1 K X

,1

ð67Þ

Ak bk r1k ðtÞ

k¼2

then the phenomenon usually known as ‘near-far effect’ takes place. As long as the previous ratio is constant, and if the user of interest turns up the power of his signal, the bit error rate will grow as a function of this increase of power. The formulation presented above allows us to probe into two possible techniques to overcome the near-far effect: † control of the power of all users, that is, power control techniques; † control of the term r1k ðtÞ that is responsible for the multi-user interference.

The different techniques developed to achieve this goal are the so-called multi-user detection techniques.

130

Broadband Wireless Mobile: 3G and Beyond

It does not matter if the spreading sequences used in the spread spectrum system are mutually orthogonal. The mobile channel is usually characterized by its multipath interference that destroys the orthogonality of such codes, thus allowing an analysis analogous to the one presented above. Advantages of multi-user detection One of the main advantages of multi-user detection techniques over power control techniques is that while the latter are to be included in the design of the system, multi-user detection techniques can be implemented independently of the spread spectrum system. The only limitation for multi-user detection to be used in a system is that the periodicity of the code should be that of the symbol or (at worse) of a few symbols. This limitation is one of complexity of the computation and the amount of memory needed to implement a Multiuser Detector (MUD), not one intrinsic to the technique. While power control techniques have to be very accurate in the refresh of the parameters used to equalize the power of the different users and techniques to employ an efficient centralized power control, which are difficult to achieve, multi-user detection techniques are dependent only on the capabilities of the receiver, i.e. the other users are not affected if one of them (or the base station) uses MUD. Neither the emitted power nor the spreading sequences are to be changed. Moreover, the ‘ceiling’ of power control techniques is that of the perfect balance of the power of all the users present in the system whereas the theoretical ‘ceiling’ of MUD is the complete removal of the influence of other users, not only the mitigation of the near-far effect. Types of multi-user detectors The different types of MUDs actually being researched or implemented is an ever growing number. Differences in performance and complexity are the reasons for the choice among other things. The classification (although far from complete and exhaustive) is as follows. † Optimum: almost at the same time that the theory of multi-user detection was formulated, the optimum multi-user detector was found. But as its complexity was exponential with the number of users, other strategies (i.e. suboptimum detectors) had to be researched. An example of this exponential complexity is that implementing an optimum MUD in a chip with up-to-date techniques (direct mapping, 0.25 m CMOS), for 5 users a power of 0.33 W and a die area of 575 mm 2 would be needed. For 10 users almost 20 W of power and an area of 3.7 £ 10 4 mm 2 would be necessary. But for 15 users almost a megawatt of power and a chip with half a meter per side would be needed to implement an optimum MUD. † Linear: the sufficient statistic vector that will be used in the multi-user detection problem can be written as ~y RAb~ 1 sn~ where R denotes the normalized crosscorrelation matrix of the spreading sequences, A is a diagonal matrix with the amplitudes of the signals of the different users as components, b is a vector with the bits of each user and n is a N(0,R) random vector. The linear MUDs are characterized by the calculation of a weight vector and the multiplication of this weight vector by the sufficient statistic and the hard limit of the result, such as:

b^ k

~ Tk ~yÞ sgnðw

for the bit of the kth user. The most common of these linear detectors [63] are:

ð68Þ

131

UMTS Air Interface

Decorrelator: the weight vector is defined by the following criteria: 2 !!2 3 K X 5 ~ argmin E4 w ~T w Ak bk ~sk ~ w

ð69Þ

k¼1

is the one that corresponds to the decorrelator. It can be seen that it is a criteria that does not take into account the noise level. MMSE: the Minimum Mean Square Error (MMSE) MUD is obtained by using the MSE definition 2 !!2 3 K X T 5 ~1 ~ 1 Þ E 4 A1 b1 w ð70Þ Ak bk ~sk 1 sn~ MSEðw k¼1

and calculating the weight vector that minimizes this function: ~1 w

~ 1Þ arg min MSEðw w1

ð71Þ

This latter MUD can be demonstrated to be the optimum linear multi-user detector. This detector also converges to the decorrelator in the limit of zero noise. Non-linear: the so-called non-linear MUDs are a huge family of detectors with varying criteria for reliable detection. It can be easily demonstrated that the best of the linear detectors cannot have the same performance ‘ceiling’ as the optimum MUD. This is the main reason for constructing non-linear detectors. Interference canceller: The successive canceller [61] is based in the fact that the strongest user, when retrieving its signal, has the easiest and most confident decision to be made. Thus, the signal of the most powerful user is estimated and then removed from the incoming mixture of signals. Then we proceed with the next powerful user; we estimate its signal, remove it from the rest and so on. This process is usually known as ‘onion peeling’. The problem with this method is the stability when a bad decision is made. This miscalculation propagates through the entire decision tree and the detector goes unstable. In multistage detection, this detector [62] is implemented in order to avoid the instabilities that the previous MUD foregoes. Basically it is a balanced version of the successive canceller where the decisions about the different users are independently taken and subtracted from each decision branch. The balancing process achieves a greater stability. This process can be repeated in stages to further enhance the detection process. These stages need not to be non-linear. Alternating stages can be combined to take advantage of the different features of each MUD. Decision feedback: much in the spirit of the decision-feedback equalizers, the decisionfeedback MUD takes advantage of previous decisions to further enhance the sequential detection of the data. As the multistage detection, both linear and non-linear MUDs are used to combat multi-user interference. Multi-user detectors proposed for UMTS The main problem with multi-user detection in an UMTS environment is that UMTS is a multirate system. The fixed chip-rate and the variation of the length of the spreading sequences, thus defining a multirate environment, makes it difficult for MUD to operate. Multirate envir-

132

Broadband Wireless Mobile: 3G and Beyond

onments are such that low-rate users interfere with a large number of bits of the other higher rate users. A single-rate multi-user asynchronous environment can be easily dealt with as the problem consists of three synchronous multi-user problems (interference with previous, actual and future bits of the user of interest). With the multirate environment the problems worsen as the number of synchronous problems grows exponentially with the number of rates and users. This is not the only problem. Although highly efficient for taking advantage of the theoretical capacity of CDMA systems, multi-user detection is a computationally costly signal processing technique. In UMTS specifications, MUDs are only specified in the uplink, i.e. implemented in Node-B. This relaxes the issue of complexity and power consumption and dissipation a bit. But MUD could be implemented in the UE eventually to better exploit the available resources. Thus the research and development of low-complexity high-speed MUDs is of great interest. To address each of the problems posed above there are two possible answers: † Groupwise serial interference cancellation [61]: this multi-user detector is more than inspired by the successive interference canceller described above. The main difference lies in the grouping of the users based on their spreading rate. The users with the highest spreading rate (lowest symbol rate) are grouped together, its interference and data are estimated, final decisions are taken for the users belonging to this group and the estimated multi-user interference of this group is fed to the following group, that is the group of users with the next lowest data rate. † Adaptive Krylov subspace detectors [64]: subspace-based detection and estimation has played a main role in other areas of signal processing for communications. In these detectors, the subspace detection and estimation is done by means of a highly efficient algorithm (both in accuracy and complexity terms) called the fast subspace decomposition. This algorithm not only separates signal and noise subspaces but estimates the number of users present in the channel, a fact much overlooked in most multi-user detection works. The multi-user detection is then carried out by means of a subspace MMSE detector. The performance of this detector is superior to the Minimum Output Energy (MOE) in its blind adaptive version and requires much fewer operations so makes it a firm candidate for implementation both in Node-B and in the UE. Research is in progress to build new non-linear detectors based on this subspace decomposition.

2.6 Conclusions 3G mobile systems will be deployed during this decade all over the world and the impact on the social, economic and technical aspects of every society profiting from their advantages will be huge. The development of services relying on these systems will be the lever to the previously mentioned transformation. Thus, a growing need for knowledge on these systems will develop in the next few years. The development of open standards and the need for convergence among all the submitted standards to ITU is a fact not previously seen in other mobile communications systems. This presentation has been made with the aim of clarity and brevity. The goal has been to reflect the main items of each system, the differences and the main points of convergence. Possible enhancements to be implemented in these systems have also been presented to show the continuous evolution taking place in this field.

UMTS Air Interface

133

References and Further Reading [1] M. Zaid, ‘Personal Mobility in PCS’, IEEE Pers. Commun., pp. 12 16, 4th Quarter, 1994. [2] M. H. Calendar, ‘Future Public Land Mobile Telecommunication Systems’, IEEE Pers. Commun., pp. 18 22, 4th Quarter, 1994. [3] International Mobile Communications 2000 (IMT 2000), ITU R Recommendations M. 687 2. [4] Framework for services supported on International Mobile Telecommunications 2000 (IMT 2000), ITU R Recommendations M.816 1. [5] International Mobile Telecommunications 2000 (IMT 2000). Network architectures, ITU R Recommendation M.817. [6] Satellite operation within International Mobile Telecommunications 2000 (IMT 2000), ITU R Recommenda tion M.818 1. [7] Requirements for the radio interface(s) for International Mobile Telecommunications 2000 (IMT 2000), ITU R Recommendation M.1034 1. [8] J. Rapeli, ‘UMTS: Targets, System Concept and Standardisation in a Global Framework’, IEEE Pers. Commun., pp. 20 28, Feb. 1995. [9] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, 25.401 v3.2.0, UTRAN overall description (Release 1999), Mar. 2000. [10] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, 25.301 v3.6.0, Radio Interface Protocol Architecture (Release 1999), Mar. 2000. [11] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, 25.202 v3.1.0, Physical Layer General (Release 1999), June 2000. [12] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, 25.302 v3.4.0, Services Provided by the Physical Layer (Release 1999), Mar. 2000. [13] A. Berrou, P. Glavieux and P. Thitimajshima, ‘Near Shannon limit error correcting coding and decoding: Turbo Codes’, Proceedings of ICC’93, Geneva, Switzerland, pp.1064 1070, May 1993. [14] C. E. Shannon, ‘A mathematical theory of communications’, Bell System Tech. J., vol. 27, pp. 379 343, 623 656, Oct. 1948. [15] G. D. Forney, Jr., Concatenated Codes, Massachusetts Institute of Technology, Cambridge, MA, 1966. [16] S. Benedetto and G. Montorsi, ‘Unveiling Turbo Codes: some results on parallel concatenated coding schemes’, IEEE Trans. Inform. Theory, vol. no. 2, pp. 409 428, Mar. 1996. [17] Divsalar and F. Pollara, ‘Turbo Codes for Deep Space Communications’, TDA Prog. Rep. 42 120. Feb. 1995. [18] J. Hagenauer, P. Robertson and L. Papke, ‘Iterative (‘TURBO’) decoding of systematic convolutional codes ´ 94, 1994. with the MAP and SOVA algorithms’, Proc. ITG [19] P. Robertson, ‘Illuminating the structure of code and decoder of parallel recursive systematic (Turbo) Codes’, ´ 94, 1994. Proc. GLOBECOM [20] R. J. McEliece, D. J. C. MacKay and J. F. Cheng, ‘Turbo decoding as an instance of Pearl’s ‘Belief Propaga tion’ algorithm’, IEEE J. Selected Areas Commun., vol. 15, no. 2, Feb. 1998. [21] P. Meshkat and J. Villasenor, ‘Generalised versions of Turbo Decoding in the framework of bayesian networks and Pearl’s belief propagation algorithm’, Proc. ICC’98, 1998. [22] J. Garcı´a Frias and J. D. Villasenor ‘Turbo codes for binary Markov Channels’, Proc. ICC’98, 1998. [23] S. Le Golf, A. Glavieux, C. Berrou, ‘Turbo codes and high spectral efficiency modulation’, Proc. ICC’94, New Orleans, LA, May 1994. [24] T. M. Duman and M. Salehi, ‘New performance bounds for Turbo Codes’, IEEE Trans. Commun., vol. 46, no. 6, June, 1998. [25] I. D. Marsland and P. T. Mathiopoulus, ‘Differential detection of Turbo codes for Rayleigh fast fading chan nels’, IEEE Commun. Lett., vol. 2, no. 2, Feb. 1998. [26] I. D. Marsland and P. T. Mathiopoulus, ‘Multiple differential detection of parallel concatenated (Turbo) codes in correlated fast Rayleigh fading’, IEEE J. Selected Areas Commun., vol. 15, no. 2, Feb. 1998. [27] E. K. Hall and S. G. Wilson, ‘Design and performance analysis of Turbo codes on Rayleigh fading channels’, Proc. CISS’96, Mar. 1996. [28] P. Komulainen and K. Pehkonen, ‘Performance evaluation of super orthogonal turbo codes in AWGN and flat Rayleigh fading channels’, IEEE J. Selected Areas Commun., vol. 15, no. 2, Feb. 1998. [29] S. Benedetto, D. Divsalar, G. Montorsi and F. Pollara, ‘Soft Output Decoding Algorithms in Iterative Decoding of Turbo Codes’, TDA Prog. Rep. 42 124, Feb. 1996.

134

Broadband Wireless Mobile: 3G and Beyond

[30] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, Working Group 4. Multiplexing and Channel Coding (FDD). TS 25.212 v3.4.0. Sep. 2000. [31] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, Working Group 4. Spreading and Modulation (FDD). TS 25.213 v3.3.0. June 2000. [32] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, Working Group 4. Physical layer procedures (FDD). TS 25.214 v3.4.0. Sep. 2000. [33] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, Working Group 4. Multiplexing and Channel Coding (TDD). TS 25.222 v3.5.0. Dec. 2000. [34] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, Working Group 4. Spreading and Modulation (TDD). TS 25.223 v3.4.0. Sep. 2000. [35] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, Working Group 4. Physical Layer Procedures (TDD). TS 25.224 v3.4.0. Sep. 2000. [36] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, 25.321 v.3.6.0 (Release 1999) ‘MAC Protocol Specification’, Dec. 2000. [37] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, 25.322 v.3.5.0. (Release 1999) ‘RLC Protocol Specification’, Dec. 2000. [38] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, 25.323 v.3.3.0. (Release 1999) ‘Packet Data Convergence Protocol (PDCP) Specification’, Sep. 2000. [39] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, 25.324 v.3.3.0. (Release 1999) ‘Broadcast/Multicast Control BMC’, Dec. 2000. [40] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, 25.331 v.3.4.1. (Release 1999) ‘RRC Protocol Specification’. [41] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, 23.110 v.3.4.0. (Release 1999) ‘UMTS Access Stratum; Services and Functions’. [42] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, TR 25.928 v1.1.0, ‘1.28Mcps Functionality for UTRA TDD Physical Layer’ July 2000. [43] 3rd Generation Partnership Project, Technical Specification Group, Radio Access Network, TR 25.924 v1.0.0, ‘Opportunity Driven Multiple Access’, Dec. 1999. [44] 3rd Generation Partnership Project 2, C.S0002 A 1 v.1.0.(Release A, Addendum1) ‘Physical Layer Standard for cdma2000 Spread Spectrum Systems’, Sep. 2000. [45] 3rd Generation Partnership Project 2, C.S0003 A (Release A) ‘Medium Access Control (MAC) Standard for cdma2000 Spread Spectrum Systems’, June 2000 [46] 3rd Generation Partnership Project 2, C.S0004 A 1 v.1.0. (Release A, Addendum1), ‘Signalling Link Access Control (LAC) Specification for cdma2000 Spread Spectrum Systems’, August 2000. [47] 3rd Generation Partnership Project 2, C.S0005 A (Release A), ‘Upper Layer (Layer 3) Signalling Standard for cdma2000 Spread Spectrum Systems’, June 2000. [48] L. R. Bahl, J. Cocke, F. Jelinek and J. Raviv, ‘Optimal Decoding of Linear Codes for Minimising Symbol Error Rate’, IEEE Trans. Inform. Theory, pp. 284 287, Mar. 1974. [49] A. J. Viterbi. ‘Error Bounds for convolutional codes and an asymptotically optimum decoding algorithm’. IEEE Trans. Inform. Theory, IT 13:260 269,1967. [50] V. Tarokh, N. Seshadri and A. R. Calderbank, ‘Space Time for High Data Rate Wireless Communication: Performance Criterion and Code Construction’. IEEE Trans. Inform. Theory, vol. 44, no. 2. [51] S. M. Alamouti, ‘A Simple Transmit Diversity Technique for Wireless Communications’, IEEE J. Selected Areas Commun. [52] V. Tarokh, H. Jafarkhani and A. R. Calderbank, ‘Space Time Block Codes from Orthogonal Designs’, IEEE Trans. Inform. Theory, vol. 45, no. 5. [53] V. Tarokh, H. Jafarkhani and A. R. Calderbank, ‘Space Time Block Codes for Wireless Communications: Performance Results’, IEEE J. Selected Areas Commun., vol. 17, no. 3. [54] A. Wittneben, ‘Base Station Modulation Diversity for Signal SIMUL CAST’, in Proc. IEEE´VTC, pp. 505 511, May 1993. [55] N. Seshadri and J. H. Winters, ‘Two signalling schemes for improving the error performance of frequency division duplex (FDD) transmission systems using transmitter antenna diversity’, Int. J. Wireless Inform. Networks, vol. 1, no. 1, 1994. [56] G. J. Foschini, ‘Layered Space Time Architecture for Wireless Communication in a Fading Environment when Using Multi Element Antennas’, Bell Labs Tech. J., Vol. 1, pp. 41 59, Autumn 1996.

UMTS Air Interface

135

[57] G. J. Foschini and M. J. Gans, ‘On Limits of Wireless Communications in a Fading Environment when Using Multiple Antennas’, Wireless Personal Commun., vol. 6, pp. 311 335, 1998. [58] B. Widrow and S. D. Stearns, Adaptive Signal Processing, Prentice Hall, Englewood Cliffs, NJ, 1985. [59] S. P. Applebaum, ‘Adaptive Arrays’, IEEE Trans. Antennas Propagation, vol. AP 24, pp. 585 598, Sep. 1976. [60] X. Wang and H. V. Poor. ‘Blind Multi user Detection: A Subspace Approach’. IEEE Trans. Inform. Theory, vol. 44, no. 2, pp. 677 689, Mar. 1998. [61] Markku J. Juntti . ‘Performance of Multi user Detection in Multirate CDMA Systems’. Wireless Pers Commun., vol. 11, no. 3, pp. 293 311, 1999. [62] H. Va¨a¨ta¨ja¨, M. Juntti and P. Kuosmanen, ‘Performance of multi user detection in TD CDMA uplink’. EUSIPCO’2000, Helsinki. [63] S. Verdu´, Multi user Detection, Cambridge University Press, 1998. [64] A. J. Caaman˜o, D. Segovia and J. Ramos, ‘Blind Adaptive Krylov Subspace multi user Detector’, unpublished paper, 2001.

3 Network Architecture 3.1 Introduction The successful deployment and worldwide acceptance of the second-generation (2G) mobile telecommunication systems combined with the need for more advanced and ubiquitous mobile services have paved the way to the initiative of the so-called third-generation (3G) mobile telecommunication systems. In this chapter, we will discuss the network architecture of 3G systems, i.e. we will describe how the network is built, what functional elements exist, what overall functionality is provided, etc. Through this discussion, the main differences between the architecture of 2G and 3G systems will be illustrated and the advanced features of 3G systems will become apparent. Finally, we will discuss the evolution of 3G systems, which shows how the 3G systems are expected to evolve in the future and lead to the next generation of mobile telecommunications systems. It is important to keep in mind that the architecture of 3G systems was based (1) on a number of market requirements and (2) on the characteristics of the installed infrastructure base pertaining to 2G systems. Indeed, since many network vendors had already invested in a large number of network elements, it was desired to keep those elements in 3G systems wherever possible. Also, both end-users and network operators have formulated a list of requirements for 3G systems. From the end-user’s point of view, the key requirements included worldwide operation, advanced services (with emphasis on multimedia applications), intelligent terminals and enhanced quality. On the other hand, from the network operator’s viewpoint, key requirements included whatever was related to increased revenue and flexible management and operation, such as, enhanced network capacity (e.g. serve more customers in a given area), increased resources utilisation, advanced network management, enhanced security, flexible and fast service deployment, etc. It is also important to point out that the network architecture was not designed having in mind the provision of telephone services but it was rather designed with multimedia services in mind. The support of multimedia services in a mobile environment was one of the primary targets and it is considered a key feature of 3G systems. Moreover, mobile access to the Internet services was another key feature. This feature was considered important because the recent evolution of Internet and its remarkable popularity called for the migration of Internet services to the mobile environment. This migration accounts for the convergence between the mobile telephony and the Internet. In this context, 3G systems are not merely mobile telephone systems similar to their 2G

138

Broadband Wireless Mobile: 3G and Beyond

counterparts, but they look more like multimedia systems with enhanced capabilities and Internet access.

3.1.1 Requirements for 3G Systems Some of the most important design requirements for 3G systems include the following: † global roaming; † wide range of operating environments, including indoor, low mobility, full mobility, and fixed wireless; † wide performance range, from voice and low speed data to very high speed packet and circuit data services; † wide range of advanced services, including voice only, simultaneous voice and data, data only, and location services; † advanced multimedia capabilities supporting multiple concurrent voice, high speed packet data, and high speed circuit data services along with sophisticated Quality of Service (QoS) management capabilities; † modular structure to support existing and future Upper Layer Signalling protocols; † seamless interoperability and handoff with existing 2G systems; † smooth evolution from existing 2G systems; † highly optimised and efficient deployments in clear spectrum, including cellular, PCS, and IMT-2000 spectrum; † support for existing 2G services, including speech coders, data services, fax services, SMS, etc.

The qualitative and quantitative differentiation between 3G and 2G systems is considered as an important factor to the success of 3G systems. For this differentiation, the following service requirements are considered very important: † † † †

significantly higher voice quality; wide range of voice and non-voice services including packet data and multimedia services; high power efficiency, especially in the mobile; efficient spectrum utilisation (which is mandatory for the provision of high data rate services); † wide range of user density and coverage; etc.

3.1.2 International Standardisation Activities The international standardisation activities for 3G systems have been mainly concentrated in the following bodies/regions: † European Telecommunications Standards Institute (ETSI) Special Mobile Group (SMG) in Europe. † Research Institute of Telecommunications Transmission (RITT) in China. † Association of Radio Industry and Business (ARIB) and Telecommunication Technology Committee (TTC) in Japan. † Telecommunications Technologies Association (TTA) in Korea. † Telecommunications Industry Association (TIA) and T1P1 in North America.

Network Architecture

139

The second-generation systems, mainly dominated by GSM, IS-136, IS-95 and PDC, have been taken into account by the related regional bodies in the design of 3G systems. This was mainly driven by the need for backward compatibility. As a result, two types of 3G core networks were standardised: one based on GSM MAP signalling and another based on IS-41 signalling. The first is being standardised by the 3rd Generation Partnership Project (3GPP), whereas the second by the 3rd Generation Partnership Project 2 (3GPP2). In general, the 3GPP initiative harmonises and standardises the similar 3G proposals proposed by ETSI, ARIB, TCC, TIA, and T1P1. The radio access of the 3GPP system is based on WCDMA and the core network is an evolution of GSM core network, based on MAP. On the other hand, 3GPP2 harmonises and standardised the 3G proposals proposed by TIA and TTA. The radio access of the 3GPP2 system is based on the so-called cdma2000 and the core network is an evolution of IS-41 core network. In addition, major international operators have initiated a harmonisation process between 3GPP and 3GPP2 in the context of the ITU-R process, which aims to result in a globally harmonised technology. In the rest of this chapter, we will focus on the 3G system standardised by 3GPP, referred to as Universal Mobile Telecommunications System (UMTS). However, we have to keep in mind that the 3G system standardised by 3GPP2 features quite similar capabilities and it is designed based on similar requirements. UMTS has been standardised in several releases, starting from Release 1999 (R99) and moving forward to Release 4 (Rel-4), Release 5 (Rel5), Release 6 (Rel-6) etc. The features of each individual release are discussed later on. UMTS is continuously evolving and up-to-date information about the individual releases can be found in the official web site of 3GPP, www.3gpp.org. Typically, one new release is frozen each year R99 was frozen in March 2000, Rel-4 was frozen in March 2001, Rel-5 was frozen in March 2002 and Rel-6 is currently being standardised. Before we get into the details of UMTS, let us briefly refer to the worldwide 3G proposals that were developed in the context of IMT-2000 program [1]. The standardisation process within ETSI started at the end of 1996. ETSI Special Mobile Group decided in January 1998 that the radio access scheme would be based on wideband CDMA (WCDMA) in paired bands (FDD), and on time-division CDMA (TD-CDMA) in unpaired bands (TDD). This radio access, called UMTS Terrestrial Radio Access (UTRA), fits into 2 £ 5 MHz spectrum allocation. ETSI has submitted the UTRA proposal to ITU-R in the context of IMT-2000 concept. Other IMT-2000 proposals from other standardisation bodies have also been submitted. China presented the ITU-R a TD-SCDMA proposal based on a synchronous TD-CDMA scheme for TDD and wireless local loop (WLL) applications. The Japanese standardisation body ARIB decided to propose a WCDMA system, aligned with the European WCDMA FDD proposal. TTA in Korea prepared two proposals: one similar to the ARIB WCDMA scheme and the other similar to the TIA cdma2000 approach. In the United States TIA prepared several proposals: UWC-136 (an evolution of IS-136), cdma2000 (an evolution of IS-95), and a WCDMA system called WIMS. T1P1 supported the WCDMA-NA system, which corresponds to UTRA FDD. WCDMA-NA and WIMS WCDMA have been merged into wideband packet CDMA (WP-CDMA) and all these technologies have been submitted to ITU-R.

140

Broadband Wireless Mobile: 3G and Beyond

3.1.3 General Aspects of 3G Systems It is important to note that, a 3G telecommunications system does not standardise the services themselves but rather it provides the means by which: † † † † †

users can connect to services from anywhere (whether roaming or not); billing and accounting functions are performed; the network is managed; security is provided; radio resources are managed, etc.

Services are offered to the users through a large variety of service providers and network operators. The user experience should be the same independent of the place and time. In other words, the user should perceive a virtual home environment (VHE) [2], wherein the same interface and service environment is maintained regardless of the location, i.e. regardless of the serving network and regardless of the access means. In a VHE the 3G network should be able to adapt the service provision to the particular capabilities of the terminal and access environment used in a particular time. Universal accessibility to services is established by enabling several access means to a common core network, (including fixed, mobile and satellite access) and multi-mode, multi-band terminals. As mentioned already, the key driver for 3G systems is the increasing demand for multimedia services. Demand is also increasing for access to multiple types of media, often used in various combinations. Thus 3G systems need to provide both narrow and wideband services (e.g. voice, data, graphics, pictures and video), in combination, on demand and on the move. This flexibility needs to be economically delivered, with costs understood by the user. It is also important to point out that, 3G systems aim to satisfy consumer (not only business) demands for personal mobile communications. Therefore, prices subscribers have to pay for the equipment and service usage should be kept to a minimum. This makes it necessary to provide common standards to build a widely accepted framework where: † low cost mass production for the manufacturers is made possible † open interfaces for the network operators, service and content providers are clearly defined.

This framework should be global in order to allow the user easy service access all over the world and in both public and private networks. 3G systems will therefore offer ubiquitous services. 3G systems aim also to provide different kinds of mobility, typically, terminal mobility, personal mobility and service mobility. Terminal mobility is provided when a user is served while on the move, regardless of network boundaries. Personal mobility is provided when a user is not restricted to a special terminal when wanting to access his or her services. This kind of mobility is usually offered by means of a common smart card technology and the provision of the virtual home environment. Service mobility is provided when a user can access his or her personalised services independently of the terminal and serving network.

141

Network Architecture

3.1.4 Chapter Outline In this chapter we will discuss the architecture of the 3G system standardised by 3GPP. As mentioned before, this 3G system is typically referred to as Universal Mobile Telecommunications System (UMTS). Section 3.2 discusses the generic network model of UMTS and defines the various highlevel domains, such as the User Equipment domain, the Access Network domain, the Core Network domain, etc. It also discusses the functional strata, including the transport stratum, the serving stratum, etc. Section 3.3 focuses on the network architecture of several UMTS releases and their particular features. In this context, it defines the various functional entities of a UMTS network, the various network interfaces and it discusses some fundamental differences with the GSM mobile networks. Section 3.4 thoroughly discusses the architecture of UMTS Terrestrial Radio Access Network (UTRAN), focusing on the internal UTRAN interfaces and the key functionality provided by UTRAN. Finally, section 3.5 discusses the network access security of UMTS, that is, the particular means provided to limit the access to network services and resources to authorised users, to encrypt the sensitive pieces of information, to verify the integrity of critical data, etc. Such functions are considered key in the deployment of UMTS. The high-level organisation of a UMTS network considered throughout this chapter is illustrated in Figure 3.1. Note that, the network is divided into the User Equipment domain, the Access Network domain, the Core Network domain and the Service domain. The core network is an evolved version of the GSM core and it is backwards compatible with GSM networks. Users access the services through the core network and by means of a particular

Figure 3.1

High level domains in a UMTS network.

142

Broadband Wireless Mobile: 3G and Beyond

access network. In UMTS, the core network is designed to be decoupled from the particular aspects of the access network and for this reason several different technologies can be used in the access network domain. Such network access technologies include the UMTS Terrestrial Radio Access (UTRAN), the standard GSM Base Station Subsystem (GSM BSS), the GSM/ EDGE Radio Access Network (GERAN), which is an evolved version of GSM BSS, the UMTS Satellite Radio Access Network (USRAN), various Broadband Radio Access Networks (BRAN), e.g. HIPERLAN/2, 802.11, etc. and also fixed access networks. At first stages of deployment, UTRAN and GSM BSS will dominate and will offer high interoperability between each other. In later stages, however, more access network options are expected to be interconnected to the UMTS core and progressively offer more access means to a common set of advanced and globally available services.

3.2 Generic Network Model The generic network architecture is a high-level representation that identifies the functional model and the physical model of the system.

3.2.1 Physical Model The physical model of the network provides a high-level physical configuration of the network. This configuration is represented by a number of physical domains connected to each other with a specific way. The UMTS physical model comprises two high-level domains: the User Equipment (UE) domain and the Infrastructure domain. This is illustrated in Figure 3.2. The reference point between these two domains is termed as Uu. 3.2.1.1 The user equipment domain The UE domain comprises of all equipment that is operated and owned by the user. The types

Figure 3.2

UMTS physical domains.

Network Architecture

143

of the equipment as well as their functional capabilities can vary between each other. However, all UE equipment must be compatible with one or more access technologies used to access the infrastructure domain. The UE domain is subdivided into two other domains: the User Services Identity Module (USIM) domain and the Mobile Equipment (ME) domain. The reference point between them is termed as Cu. The USIM includes a removable smart card that may be used in different user equipment types and it is used to provide terminal portability and terminal personalisation. The USIM corresponds to a subscription and provides the means for the infrastructure domain to securely identify the subscriber. The ME domain typically contains the equipment that execute the user applications and the radio interfacing procedures, such as physical and data link procedures. The ME domain may be physically realised in one equipment. For instance, a mobile terminal equipped with WAP microbrowser is an ME. However, it is customary to physically separate the equipment that executes the user applications and the equipment that governs the radio interface procedures. In such cases, the former equipment is referred to as Terminal Equipment (TE), while the latter is referred to as Mobile Terminal (MT) equipment. In general, a TE may be a laptop or a personal digital assistant (PDA), which runs the user applications. The TE is independent of any mobile radio issues, such as transmission, mobility management, radio resources management, etc. All these mobile radio issues are handled by the MT, which terminates the mobile protocols and the associated procedures. 3.2.1.2 The infrastructure domain The Infrastructure domain encompasses all the network equipment needed to support the endto-end user connectivity. It is further split into the Access Network (AN) domain and the Core Network (CN) domain. The AN includes the physical entities (such as radio transceivers, base stations, etc.) that manage the resources of the AN and facilitate the user access to the CN. Ideally, the AN is independent of the CN and can implement any kind of access technology ranging from the legacy fixed local loop to wireless LAN technologies, satellite access technologies, and cellular broadband technologies. Any kind of AN can be interfaced to the CN as long as it complies with the specification of the Iu reference point. Having said that, it becomes evident that the functionality of CN is decoupled from the specific AN employed and the same CN can be reused with several different AN domains. For instance, access to the same CN could be realised through an AN based on DSL technology or through an AN based on GSM radio access technology. The CN domain includes the physical entities that facilitate end-to-end connectivity, transmission of user information and signalling and, in general, the provision of telecommunication services. To support mobile access, the CN implements mobility management procedures and location management procedures. The CN domain is further sub-divided into the Serving Network (SN) domain, the Home Network (HN) domain and the Transit Network (TN) domain. The reference points between these domains are illustrated in Figure 3.2. Strictly speaking, SN, HN and TN may be considered as different instances of the same domain. This means that, at one time, some physical entities may be considered as a SN domain and, at some other time, the same physical entities may be considered as a HN domain or a TN domain. As discussed below, the definition of a domain within the CN depends on several parameters, such as the mobile user’s location and on the mobile user’s service requests.

144

Broadband Wireless Mobile: 3G and Beyond

The SN domain is composed of all the physical entities that are directly connected to the access network. The user therefore makes use of the 3G services by directly communicating with the SN. All the service requests, including mobile terminated and mobile originated calls, are handled by the SN. In addition, the SN provides the mobility management functionality needed to accommodate the user mobility. Finally, the SN establishes and routes calls on behalf of the user and it interacts with the HN domain to take into account the home environment of the user, e.g. to learn what services the user is entitled to use, what particular subscription options have been activated, what home-based services are enabled, etc. Typically, a SN serves a limited geographical area and, thus, while a user is on the go, he may enjoy 3G services through a sequence of SN domains. Each user is associated with a HN domain. This domain is effectively the network domain where the user has a 3G subscription and where some permanent user specific data is stored. Typically, a HN domain can provide the services of a SN domain for the users located in a given geographical area. However, this may not be considered as a general rule, i.e. it is possible for a HN domain to provide no SN services. When a HN domain can also provide SN domain services, then it acts as both SN domain and HN domain for the users that have a 3G subscription in this domain and are currently located in the area served by the domain. In such cases, the SN and HN domain merge into a single domain. However, when a user is located in the area served by a domain other than his HN domain, then his SN and HN domains are physically separate. In such case, we say that the user is roaming to another domain. One important characteristic of the HN domain is that its physical location is always the same no matter where the user is located. As shown in Figure 3.2, the TN domain is the components of CN that facilitates the communication between the UE and the remote party. If, for example, the mobile user makes a call to an ISDN user, then the ISDN network, which terminates the call, acts as a TN domain. Note that, for the ISDN user, this ISDN network is effectively a SN domain. The TN may be another SN, when the remote party is a mobile user served by another SN. In addition, when a call is established between two parties served by the same SN, then no TN exists. Therefore, the TN is defined on a per call/session basis and may or may not be needed for a call/session establishment.

3.2.2 Functional Model From a functional point of view, a 3G network can be decomposed into a number of functional planes, each one providing the level of functionality needed to realise a given set of services. These planes are characterised by a hierarchical relationship: a functional plane uses the services provided by the functional plane below it and provides services that are available to the functional plane above it. The decomposition of the overall functionality into several functional planes provides for a highly structural architecture and effectively maps the highlevel functional requirements into smaller sets of clearer and more specific requirements. In turn, this facilitates the development and makes it easier to identify the functional entities required. The high-level functional architecture of a 3G network is illustrated in Figure 3.3. This figure shows three primary functional planes, which in the specifications of 3GPP are referred to as strata [1]: the application stratum, the serving stratum and the transport stratum. A specific part of the transport stratum is termed the ‘access stratum’. Figure 3.3 also shows the

145

Network Architecture

Figure 3.3

UMTS functional model.

physical network entities that interact within each functional stratum. Where the interaction between two entities is represented by a dotted line, it means that this interaction is not specific to UMTS and possibly different interaction mechanisms could be employed. The application stratum is the highest-level stratum and encompasses only two peer applications, one running at the TE and another running at the ‘remote party’. The remote party could be a network server, another TE, a value-added server operated by an authority other than the 3G operator, etc. These two applications communicate transparently through the 3G network, which means that they don’t have to deal with any specific communication issues other than the communication between each other. The functional strata below the application stratum deal with all the transport and connection management issues and effectively provide a transparent communication path between the two applications. It should be noted that, the protocols used in the application stratum are not necessarily specified in 3GPP specifications. This leaves space for a vast range of applications that could be or could not be specifically developed for 3G systems. However, 3GPP specifications include also some application layer protocols, such as the USIM Application Toolkit (USAT), the Mobile Execution Environment (MExE), etc. In general, a remote application should authenticate a user before allowing him to utilise the application services and it could also provide for application level data confidentiality. Such security mechanisms are of considerable importance in the application stratum. Application-level security mechanisms are needed because the lower functional strata may not guarantee end-to-end security provision. Lack of end-to-end security could be envisioned when, for instance, the remote party is accessible through the Internet. The serving stratum is mainly used to provide access to services. Through the serving stratum a user (or an application) may request to have access to specific services. The serving stratum aims at providing the user with the requested services, or with the services that are accepTable 3.to him. In general, the serving stratum is a control stratum that received service requests from the application stratum and then configures the transport stratum to provide accepTable 3.transport (or bearer) services. For example, protocols that functionally belong to the serving stratum include the call control protocols. In particular, the serving stratum includes protocols across the following interfaces (see Figure 3.3):

146

Broadband Wireless Mobile: 3G and Beyond

† TE MT: These protocols support exchange of control information to enable the TE to request specific services. † MT SN: These protocols allow the MT to request access to services provided by the serving network domain. † USIM MT (not shown in Figure 3.3): These protocols support access to subscriberspecific information for support of functions in the UE domain.

As illustrated in Figure 3.3, there could also be serving-stratum protocols across the SN TN interface and across the TN Remote party interface; however, these protocols may not be specific to 3G architecture. As opposed to the serving stratum, which effectively deals with signalling, the transport stratum aims at providing the correct transport mechanisms to transport the actual user data between various network interfaces. The transport mechanisms across each interface are tailored to deal with the specific characteristics of that interface. For instance, the transport mechanisms across the radio interface need to cope with the radio transmission issues, such as, efficient modulation, power control, interference cancellation, etc. Obviously, such issues do not exist across other interfaces, e.g. between the SN and the TN. As identified in Figure 3.3, the 3GPP specifications specify transport mechanisms applicable only between the MT and the AN, and between the AN and the SN. For providing the required transport functionality, the transport stratum includes mechanisms such as: † mechanisms for error correction and recovery; † mechanisms to encrypt data; † mechanisms for adaptation of data to fit into the transmission format supported by the transport resources (e.g. adaptation of 13 kbps data to be transferred into a 64 kbps transport channel); † mechanisms for data transcoding to support interworking between entities using different data encoding formats.

The transport stratum in a 3G system, which aims at supporting multimedia services, should be capable of providing a vast range transport channels, each one featuring a different set of communications characteristics. The part of the transport stratum that is specific to the AN technology is called the Access Stratum (see Figure 3.3). This stratum provides services related to the transmission of data over the radio interface and the management of the radio interface. In particular, the protocols across the MT AN interface support the transfer of detailed radio-related information to coordinate the use of radio resources, and the protocols across the AN SN interface provide access of the SN to the resources of the access network. The latter interface is independent of the specific structure of the access network.

3.3 Network Architecture In this section, we present the network architecture of the 3G system specified in the 3GPP specifications [10]. The development of these specifications follow a phased approach, that is, specifications are being developed in phases, commonly referred to as releases. As we have seen in section 3.1.2, the first 3GPP release is known as Release 1999 (R99), the second as Release 4 (Rel-4), the third as Release 5 (Rel-5), etc. In every new release a list of new

Network Architecture

147

features and capabilities are supported and in addition a list of corrections to previous releases is introduced. This way, the 3GPP system is being evolved in small steps, starting from the GSM system and moving towards an advanced 3G-and-beyond system featuring all-IP transport, broadband radio access, sophisticated security (see section 3.5), voice-over-IP (VoIP), enhanced mobility, innovative services, etc. At this point, it is instructive to point out the difference between a UMTS system and a 3GPP system. A 3GPP system is typically composed of an evolved GSM core network that can be connected to several radio access networks (such as the UTRAN, the GSM radio access network, and other fixed or wireless access networks) through the standardised interfaces A, Gb and Iu. A UMTS system is similar but it does not support the A and Gb interfaces, which are required to support the GSM radio access network. As shown in Figure 3.2, a UMTS system supports only the so-called Iu interface between the core network and the access network. Therefore, a 3GPP system can be considered as an integration of a UMTS system and a GSM system. This integration is critical, as we have to provide a high degree of interworking between these two systems. To fulfil the UMTS/GSM interworking requirements, the 3GPP system features a single core network with Iu, A and Gb interfaces to the access network. The 3GPP specifications, apart from specifying the UMTS system and the evolved GSM system, they also specify the means that assure sufficient interworking between UMTS and GSM, and between the different 3GPP releases. In the rest of this section we will discuss the main aspects of 3GPP R99, Rel-4 and Rel-5.

3.3.1 3GPP Release 99 One of the key characteristics of 3GPP architecture is that the core network is an evolved GSM/GPRS core network, backwards compatible with the earlier GSM/GPRS releases. This makes sure that legacy mobiles, compliant to earlier GSM releases, are able to interwork with the 3GPP R99 network in a seamless fashion. However, the 3GPP R99 network supports a set of new features and enables the provision of more advanced (3G) services over the UTRAN radio access network. In 3GPP R99, as well as in all later 3GPP releases, the core network is functionally decomposed into two domains: the circuit-switched (CS) domain and the packet-switched (PS) domain. (Note that the PS domain provides PS-type of services, which are also called as GPRS services. In many 3GPP specifications the terms PS domain and GPRS are used interchangeably.) The CS domain is composed of the elements and interfaces that collectively provide circuit-switched type of services, i.e. services based on the reservation of dedicated circuits. This is also the kind of services provided by a GSM system. On the other hand, the PS domain is composed of the elements and interfaces that collectively provide packet-switched type of services, i.e. services that utilise resources on a demand basis. Basically, the CS domain provides ‘circuits’ (i.e. dedicated resources) for connectivity to the services, whereas the PS domain provides shared resources for connectivity to services. These shared resources are statistically multiplexed (on a demand basis) across a number of users and typically they are utilised far more efficiently when users’ traffic is bursty and uncorrelated to each other. In general therefore the PS domain is more efficient for data (bursty) communications, like web browsing, e-mail, etc., whereas the CS domain is more efficient for isochronous (i.e. constant bit-rate) type of traffic, like voice transmission. However, since most of the modern media

148

Broadband Wireless Mobile: 3G and Beyond

encoders today perform variable bit-rate encoding, it becomes evident that PS domain is not only efficient for data traffic but also for multimedia traffic. This fact accounts for the significant interest that has been developed for carrying multimedia services (including voice and video) over the PS domain. Unavoidably, the traffic carried over the CS domain progressively decreases in favour of the traffic carried over the PS domain. As a result, it is necessary to migrate to all-IP networks, which will provide only packet-switched type of services and deploy IP as a cost-effective protocol for end-to-end routing and security provision (with IPsec). If we carefully study the evolution of 3GPP and 3GPP2 specifications, we will identify that this need is acknowledged and that the network progressively migrates to an all-IP architecture. In the rest of this section we undertake a concise overview of 3GPP R99 and we discuss the basic network architecture and the key network elements and interfaces. 3.3.1.1 Basic reference architecture of 3GPP release 99 Figure 3.4 illustrates the basic reference architecture of the 3GPP R99 network, concentrating in particular to the architecture of the core network (CN). As shown, there are two types of radio access networks that can be interconnected to the 3GPP CN: the typical GMS BSS and the UTRAN, which is expensively discussed in section 3.4.1. These radio access networks are connected to the core network via standardised interfaces. In particular, the GSM BSS is connected to the CS domain via the A interface (specified in 3GPP TS 08.06 and 3GPP TS 08.08) and to the PS domain via the classical Gb interface (specified in TS 08.16 and TS 08.18). Note that all 3GPP technical specifications can be found online at [10]. On the other hand, UTRAN is connected to CS domain via the Iu-cs interface and to the PS domain via the Iu-ps interface. Both Iu-cs and Iu-ps will be discussed in section 3.4.6. The mobile station (MS) illustrated in Figure 3.4 represents the physical equipment used by a mobile network subscriber to access the 3GPP network. The MS uses the so-called Um radio interface to access the GSM BSS and the Uu interface (discussed in Chapter 2) to access UTRAN. Dual mode (GSM/UMTS) mobile stations have the capability to use both Um and Uu. The MS comprises of the Mobile Equipment (ME) and a smart card, formally called Universal Integrated Circuit Card (UICC) that contains a Subscriber Identity Module (SIM) application and/or a UMTS Subscriber Identity Module (USIM) application. The ME comprises of the Mobile Terminal (MT), which, in turn, may be decomposed into a Terminal Adapter (TA) and Terminal Equipment (TE). When the MS contains a USIM application it is also referred to as User Equipment (UE) in some of the 3GPP specifications. More information about the MS functional split and UICC can be found in TS 23.002 [3] and TS 31.120 [48] respectively. In the following sections we discuss some of the key aspects of the 3GPP R99 core network. First, we briefly discuss the key network elements and network interfaces illustrated in Figure 3.4. Later we discuss the most important features of the 3GPP R99 network. The UTRAN architecture is discussed in section 3.4.1. 3.3.1.2 Core network elements The Authentication Center (AuC) is an entity, which stores authentication data for each mobile subscriber that is used to authenticate the subscriber (more correctly, the USIM card

Network Architecture

149

Figure 3.4 The basic reference architecture of 3GPP R99.

of the subscriber) and to cipher the communication over the radio path between the mobile station and the network. The AuC interfaces only to the HLR through the H interface. When a network element, such as an SGSN or an MSC, needs to authenticate a subscriber, it requests authentication records from the HLR, which in turn contacts AuC to retrieve those authentication records. AuC stores a secret key for each mobile subscriber, commonly denoted as K. This very same key is also stored in the USIM and secured from unauthorised access. The secret key is used to generate: † authentication data which are used to authenticate the subscriber, and † Ciphering and integrity keys, which after successful authentication, are used to encrypt the data on the radio interface and to authenticate the signalling messages between the mobile and the network.

The Equipment Identity Register (EIR) in the 3GPP system is the logical entity, which is responsible for storing the International Mobile Equipment Identities (IMEIs), i.e. the unique identities of the mobile equipments used to access the network. These identities are classified as ‘white listed’, ‘grey listed’, ‘black listed’ or ‘unknown’. Typically, the network requests the mobile equipment to send its IMEI at the beginning of every phone call. If the received IMEI is included in the black list (because for instance it has been reported as stolen), then the network denies the grand of mobile services to that mobile equipment. The EIR contains one or several databases, which store the IMEIs used in the UMTS system.

150

Broadband Wireless Mobile: 3G and Beyond

The Home Location Register (HLR) is a functional element that basically stores the subscription records of the mobile subscribers (in this respect, it serves as a simple database) and also aids in the location management procedures. Depending on the number of mobile subscribers and on the capacity of the individual HLR equipments, a mobile network may contain one or more HLRs. For each subscriber, the most important information typically kept in the HLR includes: † Subscription information, containing the services the subscriber is allowed to use. † Information about the current location of the subscriber. This information enables the charging and routing of calls/packets towards the MSC and/or the SGSN that currently serves the subscriber. † The privacy exception list for location services, which indicates the privacy class of the subscriber. † A list of Gateway Mobile Location Centre (GMLC), which is also used for location services (see TS 23.071 [53]).

The HLR also stores several identifiers assigned to the user during subscription. Such identifiers include: † the International Mobile Station Identity (IMSI), i.e. the subscriber’s private identity; † one or more Mobile Station International ISDN number(s) (MSISDN), i.e. the subscriber’s public identity, or directory number; † zero or more Packet Data Protocol (PDP) address(es), e.g. statically assigned IP addresses; † the Location Management Unit (LMU) indicator (see TS 23.071).

The same IMSI or MSISDN is never assigned to more than one subscription therefore either IMSI or MSIDSN can be used to uniquely identify a subscriber to the HLR. The HLR can typically contain additional information such as: † information about subscribed teleservices and bearer services; † information about service restrictions (e.g. location areas wherein the subscriber in not allowed to roam); † a list of group IDs a subscriber is entitled to use to establish voice group or broadcast calls; † information about supplementary services; † information indicating if a GGSN in the visited network is allowed to dynamically allocate PDP addresses to the subscriber.

The organisation of subscription data is specified in 3GPP TS 23.008 and the management of subscriber data is specified in 3GPP TS 23.016. The Visitor Location Register (VLR) is the element that manages and stores information about the mobile stations roaming in an MSC area, i.e. in an area wherein a particular MSC provides services. An MSC area is typically composed of one or more location areas. In turn, every location area can be a one or more cell. Each time an MS enters a new location area it registers on that location area. The MSC serving this area informs the VLR about the new location area of the MS and in turn the VLR contacts the HLR to retrieve subscription information and determine if the subscriber is allowed to roam in the new location area and what type of services he may receive. Although a VLR may control more than one MSC areas, it usually controls only one because it is incorporated with the MSC into combined element referred to as MSC/VLR.

Network Architecture

151

The VLR keeps information needed to handle incoming calls towards the MSs located in its MSC area. The most common information includes: † the International Mobile Subscriber Identity (IMSI); † the Mobile Station International ISDN number (MSISDN); † the Mobile Station Roaming Number (MSRN), i.e. a temporary directory number assigned to a mobile for routing incoming calls with common ISUP signalling (see 3GPP TS 23.003); † the Temporary Mobile Station Identity (TMSI), see section 3.5.5 User Identity Confidentiality; † the Local Mobile Station Identity (LMSI); † the location area where the mobile station is currently located; † the identity of the SGSN that currently serves the mobile (this is applicable only when the Gs interface is supported between the SGSN and the MSC/VLR).

The VLR also contains supplementary service parameters associated to the mobile subscriber and received from the HLR. The Mobile Switching Center (MSC) constitutes the interface between the radio system and the fixed networks. The MSC performs all necessary functions in order to handle the circuit switched services to and from the mobile stations. In particular, every MSC controls a number of mobile stations (located in its MSC area) and performs the control-plane and userplane functions required for the provision of mobile CS services. The control-plane functions include call control and mobility management functions and the transport of related signalling is typically implemented with SS7. The user plane functions include functions that support the user data transmission such as transcoding, echo cancellation, etc. In 3GPP R99 the user-plane transport is usually implemented with AAL2/ATM technology. The MSC includes an Interworking Function (IWF) that provides the required functionality for interworking with fixed networks such as ISDN, PSTN and PDNs (Packet Data Networks). The IWF is required to convert the protocols used in the 3GPP CS domain to the corresponding protocols used in the external fixed network. If (for a particular type of service) these protocols are compatible, then the IWF may be bypassed. The main difference between an MSC and a typical switch in a fixed network is that the MSC performs additional functions such as functions for radio resource allocation and for mobility management. For providing service mobility, the MSC supports procedures for location registration and procedures for handover (which transfers an ongoing call from one BSS to another, or from one MSC to another). Typically, each MSC has interfaces to several base station subsystems (BSSs) and controls the MSs located in the entire radio coverage area of those BSSs (the so-called MSC area). The Gateway MSC (GMSC) is a network element used for supporting incoming calls from an external network to the CS domain, when the external network (for instance a PSTN) cannot interrogate the HLR, because, for example, it does not support the MAP protocol. In such situations, the GMSC serves as a ‘default’ gateway that receives all requests for incoming calls. The GMSC interrogates the appropriate HLR to find the addressed user’s whereabouts and then routes the call to the MSC, which currently serves this user. Usually, a mobile operator can configure a typical MSC to serve as a GMSC too. It is noted that, if the incoming call is a voice group/broadcast call, the GMSC routes the

152

Broadband Wireless Mobile: 3G and Beyond

call directly to the VGCS/VBS anchor MSC, based on the information contained in the dialled number. The Gateway GPRS Support Node (GGSN) is a network element in the PS domain that serves as a gateway providing connectivity to external packet data networks (PDNs) over the Gi interface. It could be considered as a typical IP router implementing additional functionality for supporting mobile services. Such additional functionality includes the GPRS Tunnelling Protocol (GTP), which manages user mobility by establishing and dynamically updating GTP tunnels within the PS domain. The GGSN provides an anchor point that remains fixed throughout a PS session and hides the mobility characteristics of the users. It implements also routing functionality and screening, and may optionally communicate with the HLR via the Gc interface. More detailed information on the GGSN functionality and on GTP protocol is provided in section 3.3.4 and in [57,58]. The Serving GPRS Support Node (SGSN) is a key network element in the PS domain that provides the PS-related control- and user-plane functions. In the control plane, these functions include the GPRS mobility management (GMM) and the session management (SM) functions. The SGSN supports PS services over the Gb interface and/or the Iu interface. The SGSN stores two types of subscriber data for handling originating and terminating packet data transfers: the GPRS Mobility Management (GMM) information and the Session Management (SM) information. When a mobile attaches to PS domain (i.e. to one SGSN), the SGSN establishes a GPRS Mobility Management (GMM) for that mobile. In addition, when a mobile establishes a new PS bearer, the SGSN creates a new PDP Context (see section 3.3.4). The GMM context includes the mobile’s permanent identity (IMSI), its temporary identity for the PS domain (called Packet-Temporary Mobile Station Identity P-TMSI), its current location, the address of the VLR currently serving the MS, information for authenticating the MS, etc. On the other hand, the PDP context includes information about an established PS bearer, such as the PDP address, the PDP Type, the external PDN related to this PS bearer, etc. The SGSN may optionally communicate with an MSC/VLR over the Gs interface. In such situations, the SGSN may receive paging requests from the MSC/VLR, update the MSC/VLR with a mobile’s current location, etc. More detailed information about the functionality of the Gs interface can be found in 3GPP TS 29.018. The SGSN may also interface with a Service Control Funtion (SCF) for providing service control with the CAMEL protocol (see 3GPP TS 23.078). In such cases, the PS signalling procedures may be suspended for some time and trigger the execution of CAMEL procedures. Depending on the results of these CAMEL procedures, the suspended PS signalling procedures may be resumed as normally (e.g. permit a subscriber to attach) or may be terminated (e.g. prohibit a subscriber to attach). CAMEL is typically used to provide prepaid services. The SGSN and GGSN functionalities may be implemented in the same physical node, or in different physical nodes. The SGSN and the GGSN contain IP or other Routing functionality and they may be interconnected with IP routers. When the SGSN and the GGSN are in different mobile networks, they are interconnected via the Gp interface. This interface provides the functionality of the Gn interface, plus security functionality required for the intercommunication between different mobile networks. This security functionality is

Network Architecture

153

provided by means of the border gateway (see below) and it is based on mutual agreements between different operators. The procedures for information transfer between the SGSN, the GGSN, the VLR and the HLR are defined in 3GPP TS 23.016 and 3GPP TS 23.060 [4]. Also, further discussion on the SGSN functionality and related protocols is provided in section 3.3.4 and in [57,58]. The Border Gateway (BG) is an element that serves as a gateway between the PS domain and an external IP backbone network that is used to provide connectivity with other PS domains located in other core networks. For such interconnections, the BG provides the appropriate level of security, e.g. support of IPsec tunnels, screening, etc. The BG is required only to support PS-type of services. (Note that the BG is not shown in Figure 3.4.) 3.3.1.3 Interfaces in 3GPP Release 99 Interface between MSC and GSM BSS (A-interface) The A interface operates between an MSC and a BSS and is specified primarily in 3GPP TS 08.08. The signalling protocol on this interface is called BSS Application Part (BSSAP) and implements procedures for the provision of BSS management, call handling and mobility management. Interface between the MSC and RNS (Iu-cs interface) The interface between the MSC and its RNS is discussed in section 3.4.8. Interface between SGSN and RNS (Iu ps-interface) The interface between the SGSN and the RNS is discussed in section 3.4.8. Interface between SGSN and BSS (Gb-interface) The interface between the SGSN and the BSS supports BSSGP signalling and is further discussed in section 3.3.4. The Gb interface is defined in 3GPP TS 08.14, TS 08.16 and TS 08.18. Interface between the HLR and the MSC (C-interface) The C interface is used by the MSC in order to get subscription information from the HLR about a particular subscriber and in order to obtain routing information for a call or a short message directed to that subscriber. The C interface is also used to exchange routing information, location information, subscriber status, etc. Moreover, the C interface is used as described in 3GPP TS 23.078 to support the provision of CAMEL services. Signalling on the C interface uses the Mobile Application Part (MAP), which operates on top of Transaction Capabilities protocol as specified in 3GPP TS 29.002. Interface between the HLR and the VLR (D-interface) The D interface is used to exchange location information pertaining to a mobile station and to manage the subscriber. The VLR sends to the HLR the location area wherein a mobile station is currently located and provides it (either at location updating or at call set-up) with the roaming number for that station. On the other way, the HLR sends to the VLR all the data

154

Broadband Wireless Mobile: 3G and Beyond

needed to support the service to the mobile subscriber. When the mobile station enters an MSC area controlled by a new VLR, the HLR instructs the old VLR to delete the location registration of this mobile. Data exchange over the D interface may also occur when the mobile subscriber requires a particular service, when he wants to modify data related to his subscription or when some parameters of the subscription are modified by administrative means. Moreover, D interface is used to send the CAMEL-related subscriber data to the visited mobile network, to retrieve subscriber status and location information about the mobile subscriber or to indicate suppression of a CAMEL service, etc. Signalling on the D interface is carried out with the Mobile Application Part (MAP), which uses the services of Transaction Capabilities. Interface between MSC and VLR (B-interface) The B interface is usually an internal interface since the MSC and VLR are implemented as a single network element referred to as MSC/VLR. Therefore, signalling on this interface is not standardised. The VLR serves as a location and management database for all mobile subscribers roaming in the MSC area controlled by this VLR. Therefore, each time the MSC needs data for a given mobile station located in its MSC area, it interrogates the VLR. When a mobile station performs a location updating procedure with an MSC, the MSC updates its VLR with the new location information. In addition, when a subscriber activates a supplementary service or modifies some data related to a service, the MSC informs the HLR via the VLR and the B interface. Interface between MSCs (E-interface) The E interface is used for inter-MSC signalling procedures required for example when a mobile station moves from one MSC area to another during a call and a handover procedure has to be performed. After the handover operation, the MSCs uses the E interface to transfer Ainterface signalling as necessary. Also, the E interface is used when a short message has to be transferred from a mobile station to a Short Message Center (SC) through the MSC currently serving the mobile. In this case, the E interface carries the required signalling from the MSC currently serving the mobile to the MSC that acts as the interface to the SC. Signalling on E interface is carried out with the Mobile Application Part (MAP) as specified in 3GPP TS 29.002. Interface between MSC and EIR (F-interface) The F interface is used between MSC and EIR in order to verify the status the International Mobile Equipment Identity (IMEI) received from the mobile station. Typically, if the received IMEI is included in a ‘black list’ (because for example the equipment has been reported as stolen) the service to the mobile is denied. Signalling on this interface uses the Mobile Application Part (MAP) as specified in 3GPP TS 29.002. Interface between VLRs (G-interface) The G interface is used when a mobile subscriber moves from one MSC area to another MSC area, controlled by a different VLR. In this case, the location updating procedure is carried out. This procedure may include the retrieval of the IMSI and authentication parameters from

Network Architecture

155

the old VLR. Signalling on this interface uses the Mobile Application Part (MAP) as specified in 3GPP TS 29.002. Interface between SGSN and VLR (Gr-interface) The Gr interface is equivalent to D interface and it is used when an SGSN needs to inform the HLR about the location of a mobile station. The HLR sends to the SGSN all data required to support the service to the mobile subscriber. In addition, information transfer on the Gr interface may occur when the mobile subscriber requires a particular service, when he wants to change some data related to his subscription or when his subscription parameters are modified by administrative means. Signalling on the Gr interface uses the Mobile Application Part (MAP) as specified in 3GPP TS 29.002.? Interface between SGSN and GGSN (Gn- and Gp-interface) The Gn and Gp interfaces are used to support mobility between the SGSN and GGSN. These interfaces are further discussed in section 3.3.4. The Gn/Gp interface is defined in 3GPP TS 29.060. Interface between GGSN and HLR (Gc-interface) The Gc is an optional interface that may be used by the GGSN to retrieve information about the current location of a mobile subscriber and about the subscribed services of this subscriber. This information is needed for supporting e.g. network initiated PDP context activation (see section 3.3.4) or mobile terminating SMS delivery over the PS domain. If the GGSN implements an SS7 interface the Gr interface uses the Mobile Application Part (MAP) as specified in 3GPP TS 29.002. Otherwise, if the GGSN implements no SS7 interface, then the Gr interface may be supported by any other GSN in the same mobile network, which implements an SS7 interface and an appropriate GTP-to-MAP protocol converter. Interface between SGSN and EIR (Gf-interface) The Gf interface is equivalent to F interface and it is used between the SGSN and the EIR in order to verify the status of the IMEI retrieved from the mobile station. Typically, if the received IMEI is included in a ‘black list’ (because, for example, the equipment has been reported as stolen) the service to the mobile is denied. Signalling on this interface uses the Mobile Application Part (MAP) as specified in 3GPP TS 29.002. Interface between MSC/VLR and SGSN (Gs-interface) The Gs is an optional interface that may be used between an SGSN and an MSC/VLR. With this interface the SGSN may receive paging requests from the MSC/VLR, or it may send to the MSC/VLR location information about a particular mobile. In addition, the MSC/VLR may indicate to an SGSN that a mobile station is engaged in a service handled by the MSC. The Gs interface uses the BSSAP1 protocol and it is specified in 3GPP TS 29.016 and 3GPP TS 29.018. Interface between HLR and AuC (H-Interface) The H interface is used by the HLR when it needs to receive authentication data for a

156

Broadband Wireless Mobile: 3G and Beyond

particular mobile station (e.g. because such data has to be forwarded to an MSC or SGSN). The authentication data together with a secret key assigned to a subscriber during subscription are maintained in the AuC. The AuC uses the secret key to derive the authentication data for a subscriber and forwards this data to the HLR. The protocol used to transfer the data over this interface is not standardised.

3.3.2 3GPP Release 4 3.3.2.1 Features of 3GPP Rel-4 The key features introduced in 3GPP Rel-4 are described below. Note that all of them apply to later releases too. Support of GSM/EDGE Radio Access Network (GERAN) Apart from the legacy GSM radio access network and the UMTS terrestrial radio access network (see section 3.4), 3GPP Rel-4 supports an evolved GSM radio access network, which is termed as GERAN [54]. In Rel-4 GERAN is connected to the core network using an evolved version of A and Gb interfaces, whilst in later releases GERAN uses in addition the Iu interface. GERAN uses an EDGE (Enhanced Data rates for Global Evolution) radio interface that is based on the TDMA technology of legacy GSM but introduces a set of additional radio channels with different structure, different encoding and different modulation from the typical GSM channels. These radio channels provide for increased capacity on the radio interface. An overall description of GERAN architecture can be found in [54]. Support of heterogeneous bearer mechanisms The 3GPP CN architecture is independent of the underlying transport mechanism. Ideally, any transport technology could be used, such as STM, ATM or IP, with no impact on the architecture and upper-layer protocols. This transport independency of the core network allows a mobile operator to utilise its preferred transport mechanism or any combination of transport mechanisms. Standardized transport mechanisms Although the transport mechanisms in Rel-4 and onwards may be freely chosen by the operator, these transport mechanisms need to be standardised to allow interworking across different operators. Therefore, the transport mechanisms for bearer control, call control, and other signalling are based on standardised technologies, such as SS7 or SIGTRAN. The same hold true for the transport mechanisms in the user plane. Independence of access technology The network design ensures that a common core network can be used with multiple wireless and wireline access technologies, such as ADSL, WLAN, satellite access and any other broadband radio access technology. Support for a variety of mobile equipment The network architecture supports a range of different terminal types ranging from simple

Network Architecture

157

speech only terminals to multimedia terminals and wireless IP clients (e.g. PDA or laptop). However, not all terminals may be able to support end-to-end IP capabilities, e.g. terminals that support CS voice calls only. Core network functions are separated from radio access functions The same network should support a variety of access choices, and access technologies may evolve further. Therefore network functions such as call control, service control, etc. should remain separate from access functions and ideally should be independent of choice of access. This implies that the same CN should be able to interface with a variety of RANs. High degree of decomposition The overall functionality is decomposed into a number of independent functions. This way the evolution of the overall architecture can be easily realised by evolving the individual functional components. Those components need to be kept independent as far as possible. Examples of major functions that may evolve independently include: † † † † † † † † †

security functions; bearer control in both access and network; multimedia control for multimedia sessions; switching and routing; mobility management, session control and access security functions in the PS domain; call control, mobility management and access security functions in the CS domain; location-based services; service capabilities; service features and applications.

Mobility management is decomposed to independent functions Since mobility management has evolved to a quite complex function, it is split into several individual components. These components are kept independent in order to facilitate the mobility management and the deployment of different mechanisms. As a result, mobility management is split to: † Inter-domain mobility: location of the user in terms of the domain that is currently serving the user. † Change of Point of Attachment: location of the user in terms of the address at which the user can be found. This point of attachment depends on the type of access network used and it can be provided by a wireless or fixed network. † Radio Access Mobility (or micro-mobility): location management and management of the terminal associated with changes in RA/LA within a system, or associated with changes in cell and RNC within RA/LA.

Improved speech support in the CS domain Rel-4 features enhanced speech support, for example with Transcoder-Free Operation and advanced AMR vocoders. Therefore, speech quality and transmission efficiency in the CS

158

Broadband Wireless Mobile: 3G and Beyond

core network are improved. These improvements are applicable when a call is placed on both Iu interface (UMTS) or A interface (GSM). 3.3.2.2 Basic Reference Architecture of 3GPP Rel-4 Figure 3.5 illustrates the reference architecture of 3GPP Rel-4. As mentioned in the previous section, 3GPP Rel-4 supports an additional type of radio access network as compared to 3GPP R99 the GSM/EDGE Radio Access Network (GERAN). Also, another important difference from 3GPP R99 is that the MSC is split into two functional elements, one operating in the control plane and another operating in the user plane. The first is referred to as MSC Server and the latter is referred to as Media Gateway function (MGW). In the basic reference architecture illustrated in Figure 3.5, all functional elements are considered as implemented in different equipments. Therefore, all illustrated interfaces (or reference points) are considered as external. Detailed information about interface A can be found in specifications 3GPP TS 08.06 and 3GPP TS 08.08 and information about interfaces Iu-ps and Iu-cs can be found in the 25.4xx-series of 3GPP technical specifications (see [10]).

Figure 3.5 The basic reference architecture of 3GPP Rel 4.

Network Architecture

159

Interfaces B, C, D, E, F and G use the Mobile Application Part (MAP see [55]) to exchange the data necessary to provide the mobile service. Also a general overview of all the GPRSspecific interfaces (G-series) can be found in [4] v4.x.x. Interfaces Mc, Nb, and Nc are defined in 3GPP TS 23.205 [56] and in the 29-series of 3GPP technical specifications. Note that no protocols across the H-interface are standardised. Below we briefly discuss the network elements and the interfaces that have been introduced in Rel-4 (i.e. they do not exist in R99). A Signalling Gateway (SGW) is an element that performs signalling translation at the transport level between an SS7-based transport signalling and an IP-based transport signalling (i.e. translation between Sigtran SCTP/IP transport [45,46] and SS7 MTP transport). The SGW performs no signalling translation at the application layer (e.g. MAP, CAP, BICC, ISUP) but may have to interpret the underlying SCCP [36] or SCTP layer to ensure proper routing of the signalling. The MSC Server is a signalling element that provides the call control (CC) and mobility control functionality of the MSC. The MSC server sets up and manages the mobile originated and mobile terminated calls over the CS Domain. It terminates the user-network signalling and translates it into the relevant network-to-network signalling. The MSC Server also contains a VLR, which (see section 3.3.1) maintains subscription data and CAMEL related data. The Gateway MSC Server (GMSC Server) is the element that comprises the call control and mobility control parts of a GMSC (see section 3.3.1). The Media Gateway Function (MGW) is a function in the user plane that handles the interworking with the PSTN. The MGW typically terminates bearer channels from an external circuit-switched network and media streams from a packet-switched network (e.g. RTP/ UDP/IP streams) and it bridges these bearer channels through media conversion, bearer control and payload processing. In this context, the MGW interacts with MSC server and GMSC server for resource control by using the H.248 protocol. The MGW includes the necessary resources for supporting UMTS/GSM transport media. The following interfaces have been introduced in Rel-4. Interface between (G)MSC server and MGW (Mc-interface) The Mc interface is defined between the MSC Server and the MGW, and between the GMSC Server and the MGW. It is based on H.248/IETF Megaco standard mechanisms. The functionality across the Mc reference point supports mobile-specific functions such as SRNS relocation/handover (see section 3.4.9) and anchoring. In short the key features of Mc include the following (see [3] for Rel-4): † The Mc is fully compliance with the H.248 standard. † The Mc supports flexible connection handling, which allows support of different call models and different media processing purposes not restricted to H.323 usage. † The Mc provides an open architecture and facilitates the definition of extensions. † The Mc supports dynamic sharing of MGW physical node resources. A physical MGW can be partitioned into logically separate virtual MGWs consisting of a set of statically allocated terminations. † The Mc supports dynamic sharing of transmission resources.

160

Broadband Wireless Mobile: 3G and Beyond

Interface between MSC Server and GMSC Server (Nc-interface) Across the Nc interface the network-to-network call control procedures are performed. Such network-to-network call control procedures may include the traditional ISUP procedures (which also support bearer control), or the Bearer Independent Call Control (BICC) procedures. Different transport technologies can be used on Nc such as STM, ATM and IP. When the network-to-network call control protocol employed internally in the 3GPP core network is different from ISUP, which is used in PSTN, then a signalling gateway is required for signalling translation and interworking with PSTN. This is illustrated in Figure 3.5. Interface between MGW and MGW (Nb-interface) The Nb interface is used to support the user plane traffic between MGWs. This user plane traffic could be composed for instance of speech packets with specific encoding (e.g. the classical GSM full-rate/half-rate, AMR, etc). The user plane traffic on Nc is typically transported with either RTP/UDP/IP or AAL2. In general, 3GPP Rel-4 supports different options for user data transport and bearer control on Nb such as AAL2/Q.AAL2, STM/none, RTP/ H.245. Interface between HLR and R-SGW (Mh-interface) The Mh interface supports the exchange of mobility management and subscription data information between the HLR and R99 or pre-R99 networks. This interface is required to support network users of Rel-4 or above who are roaming in networks that implement previous releases.

3.3.3 3GPP Release 5 3.3.3.1 Key Features of 3GPP Rel-5 The main characteristics of 3GPP Rel-5 are summarised below. IP transport in the UTRAN 3GPP Rel-5 enables the usage of IP technology (as an alternative to ATM technology) for the transport of signalling and user data over Iu, Iur and Iub in the UTRAN. IP transport in UTRAN provides for efficient and cost-effective utilisation of UTRAN resources. Further information on this feature can be found in 3GPP TR 25.933. High Speed Downlink Packet Access (HSDPA) In 3GPP Rel-5 several techniques are specified, which facilitate high-speed downlink packet access in the UTRAN. These techniques include adaptive modulation and coding, hybrid ARQ and other advanced features. An overall description of these techniques can be found in 3GPP TR 25.950 and in 3GPP TR 25.855. Intra Domain Connection of RAN Nodes to Multiple CN Nodes In 3GPP Rel-5 it is recognised that the requirement to have one RNC or BSC controlled by a single MSC server or SGSN lead to certain limitations. For this reason, Rel-5 provides support for allowing the BSCs and RNCs to be shared by a number of difference MSC servers

Network Architecture

161

or SGSNs. This support increases the network performance in terms of scalability, distributing the network load amongst the serving entities, and reducing the required signalling as the user roams. Detailed information about this feature can be found in 3GPP TS 23.236. GERAN Iu interface In 3GPP Rel-5 architecture the GSM/EDGE Radio Access Network (see Figure 3.5) can be connected to the core network over the standard Iu interface of UMTS. The Iu interface is discussed in section 3.4.6. End-to-End QoS for PS Domain 3GPP Rel-5 provides a framework for end-to-end Quality of Service, i.e. for IP bearers with guaranteed QoS properties not only in the UMTS domain but also within the domain of external IP networks. For this purpose, interworking between the UMTS QoS mechanisms and the IP QoS mechanisms (defined by IETF mainly DiffServ and IntServ) are required. The architecture and the mechanism for the provision of end-to-end QoS are specified in 3GPP TS 23.207. Provisioning of IP-based Multimedia Services 3GPP Rel-5 supports a special core network subsystem, which is called the IP Multimedia CN Subsystem (IMS). IMS is the key characteristic of 3GPP Rel-5. It is virtually a signalling system on top of PS domain (i.e. accessible over typical PS bearers) that enables the provision of IP multimedia services. From the mobile point of view, IMS is a signalling application that can be used to provide various services. Note that IMS implements the signalling procedures for the provision of those services, however, it does not provide the services themselves. The services are not specified in 3GPP specifications; only the interfaces towards the users and the service platforms are specified. Therefore, IMS is as an open signalling system, based on standard Internet technology, which supports the migration of Internet applications (for example, voice-over-IP, multimedia messaging, video-conferencing, etc.) to the mobile environment and offers enhanced service control capabilities. Mobile users may access IMS through the bearer services provided by the PS domain. IMS service mobility is maintained by exploiting the mobility management services of PS domain. Note that IMS is independent from the CS domain although some network elements may be shared with the CS domain. Note also that access to IMS is not necessarily provided over the PS domain. In fact, any suitable access means can be used, such as wireless LANs, or wireline access technology. In any case, the technology that provides access to IMS needs to handle the user mobility and in particular to hide the mobility of the user when it roams within the same mobile network. This is because IMS does not provide mobility management and assumes that the user is always accessible through the same reference point (also called anchor point). To achieve access independence and to maintain a smooth interoperation with wireline terminals across the Internet, IMS needs to conform to standard IETF protocols. For this reason, the Session Initiation Protocol (SIP) [49] has been selected as the IMS signalling protocol to/from the users and to/from the service control platforms (see TS 23.218 [51]).

162

Broadband Wireless Mobile: 3G and Beyond

3.3.3.2 Basic Reference Architecture of 3GPP Rel-5 If we except IMS, the basic reference network architecture of 3GPP Rel-5 is almost identical with the basic reference network architecture of Rel-4. The basic architecture of the 3GPP Rel-5 is depicted in Figure 3.6, which focuses on the architecture of IMS. As can be seen from this figure, IMS can practically interoperate with other multimedia-capable systems, such as other multimedia IP networks (usually, featuring H.323 or MBONE technologies to enable multimedia applications). When it comes to voice applications, the IMS can be interconnected to legacy PSTN networks, or legacy mobile networks, such as the GSM, and establish end-to-end voice transactions. The IMS features all the interworking capabilities required to map between voice-over-IP and voice-over-PSTN/ISDN/GSM. Below we explore the main core network elements of 3GPP Rel-5 (virtually they are IMSspecific elements). The Call Session Control Function (CSCF) is the element that manages the IP multimedia sessions by using the SIP protocol. There are three roles defined for a CSCF element: † Proxy CSCF (P-CSCF), † Serving CSCF (S-CSCF) † Interrogating CSCF (I-CSCF).

Each CSCF element in the IMS domain can have one of those three roles. The Proxy-CSCF (P-CSCF) is the CSCF that directly communicates with the mobile terminals through the Gi interface. In other words, the P-CSCF is the first contact point in the IMS. Mobiles discover the address of one or more P-CSCF by means of the so-called P-CSCF Discovery procedure (see TS 23.228 [50]). The P-CSCF effectively serves as a SIP proxy. It

Figure 3.6

The basic reference architecture of 3GPP Rel 5.

Network Architecture

163

receives the requests issued by mobiles and, possibly after translation, it forwards them to another CSCF, which, depending on the message, is either an I-CSCF or a S-CSCF. In some cases, the P-CSCF may choose not to forward a SIP message but rather to handle it by itself and respond to the mobile (for instance, to report an error condition). The most important functions carried out by a P-CSCF are summarised as follows: † The P-CSCF forwards SIP register requests from the mobiles to the appropriate I-CSCFs, which are identified by means of the home domain name of the mobile and the DNS system. † The P-CSCF forwards SIP messages from the mobile to the appropriate S-CSCFs, which have been memorized before when the mobiles have registered to IMS services. † The P-CSCF may modify the contents of mobile originated and mobile terminated SIP messages based on its capabilities and on a set of provisioned rules defined by the network operator. † The P-CSCF generates Charging Data Records (CDR), which are forwarding to a mediation gateway and are used for billing the utilised services. † The P-CSCF identifies the requests for emergency sessions and selects the appropriate SCSCFs. † The P-CSCF maintains a security association with every registered mobile, which is generated in the context of the registration procedure. This way SIP messages between the P-CSCF and a mobile are subject to integrity protection and optionally to encryption. † The P-CSCF performs compression/decompression of SIP messages in order to save resources on the radio interface. † The P-CSCF issues authorisations for utilising PS domain resources. These authorisations are subsequently used by the mobiles when they request from the PS domain to allocate the appropriate bearer resources. These resources transport the media of the IMS sessions.

The Interrogating CSCF (I-CSCF) is the contact point of an operator’s IMS network and may be used to hide the internal configuration details of the IMS network deployed by that operator. There may be multiple I-CSCFs in an IMS network. The most important functions performed by an I-CSCF are summarised as follows: † During a mobile’s registration, the I-CSCF is responsible for assigning an S-CSCF to that mobile. This S-CSCF will subsequently serve the mobile for as long as it is registered to IMS. † The I-CSCF routes SIP messages received from an external network towards the appropriate S-CSCF (the one that has been previously selected during the registration phase). † The I-CSCF retrieves from HSS the address of the S-CSCF that has been assigned to serving a particular mobile. † The I-CSCF generates Charging Data Records (CDR), which are used for billing the utilised IMS services.

The Serving-CSCF (S-CSCF) is the element that provides the session control services. Typically, an IMS network has one or more S-CSCF depending on the number of subscribers and the capabilities required. It is noted that, different S-CSCFs may provide different functionalities. The S-CSCF that is assigned to a user must be capable of supporting the subscribed services of the user. The most important functions performed by an S-CSCF are summarised as follows:

164

Broadband Wireless Mobile: 3G and Beyond

† The S-CSCF accepts registration requests from the mobiles and after successful authentication it creates a mobile context. † The S-CSCF controls the IMS sessions of the registered mobiles. † The S-CSCF may service the received requests internally or choose to forward them to another entity, possibly an application server. † The S-CSCF may behave as a user agent and terminate SIP sessions. † The S-CSCF interfaces with the service platforms for the provision of services (e.g. voice mailbox). † Upon receiving a SIP request for a new session, the S-CSCF discovers the address of the ICSCF of the terminating IMS network and forwards the SIP request to that I-CSCF. † When a call needs to be routed to a user accessed through e.g the Internet, the S-CSCF forwards the SIP signalling to another SIP server located outside the IMS, e.g. in the Internet. † When a call needs to be routed to a user accessed through the PSTN or the CS domain, the S-CSCF forwards the SIP signalling to a Breakout Gateway Control Function (BGCF). † The S-CSCF forwards mobile terminating SIP signalling to the appropriate P-CSCF that serves the target mobile. If the target mobile is in a network that applies configuration hiding, the S-CSCF forwards mobile terminating signalling to the appropriate I-CSCF. † Route mobile terminating requests to the CS domain for subscribers who have requested to receive incoming sessions via the CS domain. † The S-CSCF generates Charging Data Records (CDR), which are forwarding to a mediation gateway and are used for billing the utilised services.

The Media Gateway Control Function (MGCF) receives incoming call requests from an external legacy (e.g. PSTN) network and based on the received routing number selects a CSCF to communicate with in order to handle the received call request. The MGCF controls the parts of the call state that pertain to connection control for media channels in an IMMGW. It also performs protocol conversion between ISUP and the SIP protocol. The Breakout Gateway Control Function (BGCF) is an element that selects the network in which the breakout towards the PSTN will occur. It also selects the appropriate MGCF in that network that will terminate the SIP signalling and connect to the PSTN. The IP Multimedia Media Gateway Function (IM-MGW) terminates bearer channels from a switched-circuit network and media streams from a packet network. It typically supports media conversion, bearer control and payload processing (e.g. echo cancellation). In this context, it interacts with the MGCF for resource control, it handles resources such as echo cancellers, etc., and it maintains a set of codecs. The Multimedia Resource Function (MRF) performs multiparty call and multimedia conferencing functions. Typically, the MRF performs the same functions of an MCU in an H.323 network. The MRF is also responsible for bearer control (with GGSN and IM-MGW) for multiparty/multimedia conferences. It may communicate with a CSCF for service validation for multiparty/multimedia sessions. Note that in the later version of 3GPP specifications, the MRF, has been split into the MRFC and the MRFP (see 3GPP TS 23.002 v5.6.0). Below we briefly discuss the new interfaces introduced in Rel-5 (they don’t exist in previous releases).

165

Network Architecture

Interface between HSS and CSCF (Cx-interface) The Cx interface is defined in 3GPP TS 29.228 and 3GPP TS 29.229 and it is based on the Diameter protocol specified by IETF. It is used to provide information transfer between a CSCF and the HSS. This information transfer is required to support the procedures related to the Serving CSCF assignment, the procedures related to transfer of routing information from HSS to CSCF, and the procedures related to MS-HSS information tunnelling via a CSCF. Interface between CSCF and MS (Gm-interface) The Gm interface composes the user-network signalling interface in IMS. (The Gm interface is not shown in Figure 3.6.) It carries SIP signalling between a mobile station and its serving Proxy CSCF and provides session management, registration and supplementary services control. More specifically, it mainly supports procedures for registration to a Serving CSCF, procedures for session origination and termination, and procedures for authentication and authorisation. Interface between MGCF and IM-MGW (Mc-interface) The Mc interface used between the MGCF and IM-MGW is the same with the interface used between the MSC Server and MGW in Rel-4 as has been described. Interface between MGCF and CSCF (Mg-interface) The Mg interface supports SIP signalling for the communication between the MGCF and CSCF. Interface between CSCF and external multimedia IP networks (Mm-interface) The Mm interface is an IP interface between a CSCF and an external IP network. It is typically used to support interworking with other multimedia protocols used in the Internet, such as the H.323. Interface between CSCF and CSCF (Mw-interface) The Mw interface operates between two CSCF elements, e.g. between a P-CSCF and a SCSCF or I-CSCF. This is the primary signalling interface in IMS and it is based on the SIP protocol. The procedures supported on Mw can be found in 3GPP TS 24.229. Interface between CSCF

BGCF (Mi-interface)

The Mi interface carries SIP signalling and allows an S-CSCF to forward a multimedia session to the Breakout Gateway Control Function and interwork with PSTN networks. Interface between BGCF

MGCF (Mj-interface)

The Mj interface allows a Breakout Gateway Control Function to forward a multimedia session coming from external PSTN network to a Media Gateway Control Function. This way, interworking with PSTN networks is supported.

166

Broadband Wireless Mobile: 3G and Beyond

3.3.4 An Overview of PS Domain Protocols In this section we describe the main protocols used for the provision of PS services in Gb mode, i.e. when the Gb interface is used between the radio access network and the core network. This description is applicable to all 3GPP releases, including those discussed in the previous sections. The specific details and the protocol architecture of the Iu interface are discussed in section 3.4.6. The protocol architecture used for the provision of PS services in Gb mode is illustrated in Figure 3.6a. Note that the PS domain mainly provides a PS bearer, i.e. a communications channel between the MS and the GGSN with specific quality of service properties. This PS bearer provides connectivity between the routing layer in the mobile and the routing layer in the GGSN. Figure 3.6a assumes that IP is used as a routing protocol, but other protocols could also be supported. This PS bearer is established with the appropriate Session Management (SM) procedures. After the successful establishment, the PS bearer is described in the MS and in the GGSN (and SGSN) with a so-called PDP Context, which is a collection of information that relates to the operation of this PS bearer. Every packet data unit between the MS and the GGSN is transferred over the appropriate PS bearer or PDP Context. Note that the terms PS bearer and PDP context are typically used interchangeably. The Subnetwork Dependent Convergence Protocol (SNDCP) runs between the MS and the SGSN, and it is specified in 3GPP TS 04.64. It is the first layer that receives the user IP datagrams for transmission. SNDCP basically provides: † † † † †

acknowledged and unacknowledged transport services; compression of TCP/IP headers; compression of user data; datagram segmentation/reassembly; PDP Context multiplexing.

The segmentation/reassembly function ensures that the length of data units sent to the LLC layer does not exceed a maximum pre-negotiated value. For example, when this maximum

Figure 3.6a Protocols used for the provision of PS services over the Gb interface.

Network Architecture

167

value is 500 octets, then IP datagrams of 1500 octets will be segmented into three SNDCP data units. Each one will be transmitted separately and reassembled by the receiving SNDCP layer. A PDP Context essentially represents a virtual connection between an MS and an external PDN. The PDP Context multiplexing is a function that: † routes each data unit received on a particular PDP Context to the appropriate upper layer and † routes each data unit arrived from an upper layer to the appropriate PDP Context.

For example, assume a situation where the MS has set up two PDP Contexts, both with type IP but with different IP addresses. One PDP Context could be linked to a remote PDN A and the other could be linked to a remote PDN B. In this case, there are two different logical interfaces at the bottom of the IP layer, one for each PDP Context. The SNDCP layer is the entity that multiplexes data to and from these two logical interfaces. The Logical Link Control (LLC) protocol runs also between the MS and the SGSN, and it is specified in 3GPP TS 04.64. LLC basically provides data link services, as specified in the Open System Interconnection (OSI) seven-layer model. In particular, LLC provides one or more (up to 16) separate logical links (LLs) between the MS and the SGSN, which are distinguished into user-LLs (used to carry user data) and control-LLs (used to carry signalling). There can be up to four user-LLs, while there are basically three control-LLs: one for exchanging GPRS Mobility Management (GMM) and Session Management (SM) signalling, another to support the Short Message Service (SMS) and a third to support Location Services (LCS) (see 3GPP TS 03.71). The user-LLs are established dynamically, in the context of the PDP Context Activation procedure [57], and their properties are negotiated between the MS and the SGSN during the establishment phase. Negotiated properties typically include: † † † †

the data transfer mode (acknowledged versus unacknowledged); the maximum length of transmission units; timer values; flow control parameters, etc.

On the other hand, the control-LLs have pre-defined properties and they are automatically set up right after the MS registers to the GPRS network. It should be noted that each user-LL carries data for one or more PDP Contexts, all sharing the same QoS. Control LLs operate only in unacknowledged mode, which basically provides an unreliable transport service. On the other hand, user-LLs operate either in unacknowledged mode or in acknowledged mode, depending on the reliability requirements. The latter mode provides reliable data transport by: † detecting and re-transmitting erroneous data units; † maintaining the sequential order of data units, and † providing flow control.

Another service provided by the LLC layer is ciphering. This service can be provided in both acknowledged and unacknowledged mode of operation and, therefore, all LLs can be secured and protected from eavesdropping. The Radio Link Control (RLC) and Medium Access Control (MAC) protocols run between

168

Broadband Wireless Mobile: 3G and Beyond

the MS and the PCU, and they are specified in 3GPP TS 04.60. The RLC provides the procedures for unacknowledged or acknowledged operation over the radio interface. It also provides segmentation and reassembly of LLC data units into fixed-size RLC/MAC blocks. In RLC acknowledged mode of operation, RLC also provides the error correction procedures enabling the selective retransmission of unsuccessfully delivered RLC/MAC blocks. Additionally, in this mode of operation, the RLC layer preserves the order of higher layer data units provided to it. Note that, while LLC provides transport services between the MS and the SGSN, the RLC provides similar transport services between the MS and the PCU. The MAC layer implements the procedures that enable multiple mobile stations to share a common radio resource, which may consist of several physical channels. In particular, in the uplink direction (MS to network), the MAC layer provides the procedures, including contention resolution, for the arbitration between multiple mobile stations, which simultaneously attempt to access the shared transmission medium. In the downlink direction (network to MS), the MAC layer provides the procedures for queuing and scheduling of access attempts. More details are provided below. The MAC function in the network maintains a list of active MSs, which are mobile stations with pending uplink transmissions. These MSs have previously requested permission to content for uplink resources and the network has responded positively to their request. Each active MS is associated with a set of committed QoS attributes, such as delay, throughput, etc. These QoS attributes were negotiated when the MS requested uplink resources. The main function of the MAC layer in the network is to implement a scheduling function (in the uplink direction), which successively assigns the common uplink resource to active MSs in a way that guarantees that each MS receives its committed QoS. A similar scheduling function is also implemented in the downlink direction. From the above, it is obvious that, every cell supporting PS services in Gb mode features a central authority, which: † arbitrates the access to common uplink resources (by providing an uplink scheduling function) and † administers the transmission on the downlink resources (by providing a downlink scheduling function).

These scheduling functions are part of the functions required to guarantee the provisioning of QoS on the radio interface, and are implementation dependent. The Base Station Subsystem GPRS Protocol (BSSGP) runs across the Gb interface, and it is specified in 3GPP TS 08.18. BSSGP basically provides: † unreliable transport of LLC data units between the PCU and the SGSN and † flow control in the downlink direction.

The flow control aims to prevent the flooding of buffers in the PCU and to conform the transmission rate on Gb (from SGSN to PCU) to the transmission rate on the radio interface (from PCU to MS). Flow control in the uplink direction is not provided because it is assumed that uplink resources on Gb are suitably dimensioned, and are significantly greater than the corresponding uplink resources on the radio interface. BSSGP provides unreliable transport because the reliability of the underlying frame relay network is considered sufficient enough to meet the required reliability level on Gb. BSSGP provides also addressing services, which are used to identify a given MS in uplink

Network Architecture

169

and downlink directions, and a particular cell. In the downlink direction, each BSSGP data unit typically carries an LLC data unit, the identity of the target MS, a set of radio-related parameters (identifying the radio capabilities of the target MS) and a set of QoS attributes needed by the MAC downlink scheduling function. The identity of the target cell is specified by means of a BSSGP Virtual Channel Identifier (BVCI), which eventually maps to a frame relay virtual channel. In the uplink direction, each BSSGP data unit typically carries an LLC data unit, the identity of the source MS, the identity of the source cell and a corresponding set of QoS attributes. The mobility management function in the SGSN uses the source cell identity to identify the cell wherein the source MS is located. As shown in Figure 3.6a, the GPRS Tunneling Protocol (GTP) runs between the SGSN and the GGSN. In general, however, GTP also runs between two SGSNs. GTP provides an unreliable transport function (usually runs on top of UDP) and a set of signalling functions primarily used for tunnel management and mobility management. The transport service of GTP is used to carry user-originated IP datagrams (or any other supported packet unit) into GTP tunnels. GTP tunnels are necessary between the SGSN and the GGSN for routing purposes [58]. They are also necessary for correlating user-originated IP datagrams to PDP Contexts. By means of this correlation, a GGSN knows how to treat an IP datagram received from an SGSN, for example to which external PDN to forward this datagram, and an SGSN knows how to treat an IP datagram received from a GGSN (or another SGSN), for example what QoS mechanisms to apply to this datagram, to which cell to forward this datagram, etc.

3.4 UMTS Terrestrial Radio Access Network As discussed before, the UMTS Terrestrial Radio Access Network (UTRAN) [14] is one of the several possible radio access networks (RANs) that could be used in a UMTS network. The main purpose of UTRAN is to facilitate the communication between the user equipment and the core network. The radio access network provides and manages the wireless resources required for the signalling and user data transmission between the user equipment and the core network. As we have already seen in the functional UMTS model, a radio access network is part of the Access Stratum and, therefore, it offers the means for the user to access the services provides by the core network. This section describes the fundamental architecture of UTRAN, as well as the most significant UTRAN functions.

3.4.1 UTRAN Architecture As illustrated in Figure 3.7, UTRAN is composed of a collection of Radio Network Subsystems (RNS), each one connected to the core network through the so-called Iu reference point. An RNS is responsible for managing the radio resources allocated to a number of cells and for the transmission/reception in these cells. Every RNS consists of one Radio Network Controller (RNC) and one or more Node-Bs. As shown in Figure 3.7, the reference point connecting two RNCs is referred to as Iur, while the reference point connecting an RNC with a Node-B is referred to as Iub. Each Node-B is controlled by only one RNC and it supports the physical wireless interface with the mobile terminal, i.e. the Uu interface. As explained in Chapter 2, the Uu interface is based on W-CDMA technology and it supports either Frequency Division Duplex (FDD) mode or Time Division Duplex (TDD) mode. Each RNC is assigned a pool of radio resources (for example frequencies, CDMA codes,

170

Broadband Wireless Mobile: 3G and Beyond

Figure 3.7

Network schematic diagram

etc.) and is responsible for managing those radio resources and allocating them to the mobile users on a demand basis. Each UE connected to UTRAN is served by a specific RNC, which is called the Serving RNC (SRNC). The SRNS controls the signalling connection between the UE and the UTRAN and it also controls the Iu signalling connection for this UE. Formally, the signalling connection between the UE and the UTRAN is called the RRC connection, because the signalling protocol between the UE and the SRNC and the RRC protocol [28]. The RNC that controls a specific set of UTRAN access points, i.e. one or more Node-Bs, serves as the Controlling RNC (CRNC) for these Node-Bs. There is only one CRNC for any Node-B and that CRNC has the overall control of the logical resources of that Node-B. In general, an RRC connection originates at the UE, passes transparently through a Node-B and its CRNC and terminates at the SRNC. The SRNC and the CRNC may or may not be implemented in the same RNC node. Figure 3.8(a) shows a situation where the CRNC of the illustrated Node-B and the SRNC of the UE correspond to the same RNC node. On the contrary, Figure 3.8(b) shows another situation where the CRNC of the illustrated Node-B and the SRNC of the UE correspond to different RNC nodes. It is clear that the role of controlling RNC makes sense only in connection with a specific Node-B, while the role of Serving RNC makes sense only in connection with a specific UE. In Figure 3.8(b) the CRNC of Node-B serves as a Drift RNC (DRNC) with respect to the UE. In general, a DRNC is the role an RNC can take with respect to a specific RRC connection

171

Network Architecture

Figure 3.8

The roles of SRNC and CRNC/DRNC.

between a UE and UTRAN. An RNC that supports the SRNC with radio resources when the connection between the UTRAN and the UE need to use cell(s) controlled by this RNC, is referred to as DRNC. In this chapter, the UTRAN Iu, Iub and Iur interfaces are collectively referred to as Ix interfaces. Below, we describe the general architectural aspects of the Ix interfaces. The architectural aspects of the Uu interface are presented in Chapter 2. 3.4.1.1 Architecture of UTRAN Ix interfaces Functionally, the architecture of UTRAN Ix interfaces is decomposed into two layers, the Radio-Network Layer (RNL) and the Transport-Network Layer (TNL). This is illustrated in Figure 3.9. The Radio-Network layer is associated with UTRAN-specific signalling and userdata protocols used between UTRAN nodes, while the Transport-Network layer is associated with the transport mechanisms used to transport information (either signalling or user data) between UTRAN nodes. Ideally, these two layers are completely independent and, therefore, the UTRAN-specific protocols can be implemented over several transport mechanisms. For instance, one operator may choose to use ATM technology in the Transport-Network layer, while another one may choose IP technology. The UTRAN Radio-Network layer has been specified to be independent of the underlying transport mechanisms and thus any transport mechanism can be used in UTRAN, as long as it satisfies the data transmission requirements (e.g. delay, throughput, etc). In first UMTS releases, the UTRAN Network-Transport layer is

172

Broadband Wireless Mobile: 3G and Beyond

Figure 3.9 The concept of transport network and radio network layers.

based on ATM technology, while in later UMTS releases it evolves to also support IP-based transport [29]. One of the key features of UTRAN is that the transport-network layer in the control-plane is independent (or logically separated) from the transport-network layer in the user-plane. This means that the transport mechanisms used to transport signalling are generally different and separate from the transport mechanisms used to transport user data. The signalling and data transport mechanisms used across the various interfaces, i.e. across Iub, Iur and Iu, will be discussed later on, when these interfaces are discussed in more detail. As already mentioned above, another key feature of UTRAN is that the radio-network layer and the transport-network layer are fully separated and independent. This effectively means that the control- and user-plane functions in UTRAN are fully separated from the underlying transport functions. Therefore, the transport functions and the UTRAN-specific control- and user-plane functions can evolve independently. 3.4.1.2 Protocol model of Ix interfaces Figure 3.10 illustrates the general model used to describe the protocol architecture of UTRAN Ix interfaces. This model is composed of a number of horizontal and vertical layers. These layers are logically independent from each other and this accounts for a highly modular and expandable architecture. Indeed, each module or layer can be replaced or evolve independently from the rest of the modules and, therefore, future requirements can be addressed be updating only the correct module(s), rather than updating the entire protocol architecture. In the horizontal direction, there is the radio-network layer and the transport-network layer. As already explained, the former entails the UTRAN-specific signalling and user data protocols, while the latter entails the underlying transport protocols that transport the UTRANspecific protocol data units. All UTRAN-specific issues are handled only by the radio network layer. The transport network layer is based on standard transport technology, such as ATM and IP. In the vertical direction there are two main planes: the control plane and the user plane,

173

Network Architecture

both extending across the radio-network layer and the transport-network layer. The control plane contains a UTRAN Signalling Protocol (RANAP, RNSAP, or NBAP, which will be addressed later) and an associated transport mechanism for the transport of the signalling messages between the UTRAN nodes. As discussed later, the UTRAN Signalling Protocols are used for setting up data transports (or data bearers) in the radio network layer. The User plane contains a UTRAN protocol that deal with user-specific data and an associated transport mechanism for the transport of the user-specific data between UTRAN nodes. Typical UTRAN protocols that deal with user-specific data (or data streams) include the Medium Access Control (MAC) protocol and Radio Link Control (RLC) protocol. As illustrated in Figure 3.10, both Signalling Transport(s) and Data Transport(s) are defined in the Transport Network User Plane because, from the transport network point of view, they both transport protocol units that pass transparently through the transport network. In other words, the transport network treats these protocol units as data units that need to be transferred without any interpretation. On the other hand, the protocols defined in the Transport Network Control Plane, are Transport-Signalling protocols (generally referred to as ALCAP protocols) that generate signalling messages for controlling the transport network. These messages are interpreted by the transport network and are mainly used to dynamically set up and release virtual channels within the transport network (of course, when the transport network supports virtual circuits). As an example, when a new conversational call needs to be established and the transport network in based on ATM, a new virtual circuit, for example an AAL2 VC, would need to be established between the Node-B and its CRNC. This AAL2 VC is created and controlled by means of the transport-signalling protocol. It should be noted that the Signalling Transport for the UTRAN Signalling Protocol(s) may or may not be of the same type as the Signalling Transport(s) for the ALCAP.

Figure 3.10

General protocol model of UTRAN interfaces.

174

Broadband Wireless Mobile: 3G and Beyond

The specific details of the protocol architecture that apply to each UTRAN interface will be discussed later on, when each UTRAN interface is discussed.

3.4.2 UTRAN Functions Below we summarise some of the most important functions provided by UTRAN. All these functions are provides by means of several elementary procedures employed in the UTRAN Ix and Uu interfaces. Some of these elementary procedures will be discussed later. 3.4.2.1 Information broadcasting In every cell, UTRAN broadcasts information needed by the mobiles to perform specific tasks, or used to collect specific system information about the network. For example, the broadcast information includes the network and the cell identities, information for location registration purposes, specific access rights applicable in a given time, information for transmission power control, information regarding the configuration of the transport channels, information for cell selection and cell-reselection, the frequency bands used, etc. In addition, physical-layer information is broadcast to aid mobile acquire synchronisation and decode the downlink control channels. 3.4.2.2 Security provision The security mechanisms applied to protect the user data and the signalling information against malicious attacks, are carried out between the mobile and the UTRAN. In particular, the UTRAN is responsible for setting up the security mode before any sensitive data is communicated. In the context of security both ciphering and integrity protection are provided. 3.4.2.3 Mobility management The main mobility management functionality required for a mobile in CONNECTED mode is provided by the UTRAN. For this mobility management the following procedures are provided: † Radio measurements: The quality of the radio environment provided to a mobile is calculated by means of specific radio measurements, taken on the current and the neighboring cells. Typical measurements include the signal strength and the estimated bit error rate of the current and the neighboring cells, the estimation of the propagation environment, the received interference level, the Doppler shift, etc. † Handover decision: From the gathered radio measurements the UTRAN estimates the quality of the current radio channel and the quality of the neighboring radio channels. Then it compares the overall quality of service provided with the current radio channel with the requested limits and with the estimated quality of service of the neighboring cells. Depending on the outcome of this comparison, the UTRAN may activate the handover procedure or the macro-diversity procedureand transfer the communication path to another radio channel in another cell. In addition, the UTRAN may activate the handover

Network Architecture

†

†

† †

175

procedure to balance the traffic loading between several radio cells. In such cases, a mobile is commanded to move to a neighboring cell, which is less loaded than the current cell and which can also provides accepTable 3.quality of service. Macro-diversity procedure: Before a mobile is handed over to another radio channel the macro-diversity mode may be activated. In this mode, the mobile is assigned an additional radio channel, possibly in another cell, to support the same call. Therefore, the information transmitted by the mobile is simultaneously transmitted on two radio channels. In addition, the downlink information is transmitted to the mobile by two different radio channels. For this macro-diversity technique to work, the UTRAN needs to carry out a macro-diversity combining/splitting function, which combines the two uplink streams into one and splits the downlink stream into two, one for each downlink radio channel. Handover procedure: This procedure is executed when the current call needs to be switched to another radio channel, which is considered more appropriate. The handover procedure is decomposed into several phases. First, a handover initiation is executed, which identifies the need for handover to the related elements. These elements will need to take some action in order to realise the handover. Then, the handover resource allocation takes place wherein the some new resources are allocated and activated to support the call after the handover. Subsequently, the handover execution is carried out wherein the mobile is commanded to switch to the new channel. When the mobile actually changes channel, the call is switched to the new path, which has already been activated during the handover resource allocation phase. Finally, the handover completiontakes place wherein the old resources, which supported the call before the handover, are released. It should be noted that, a handover could be hard or soft. A soft handover initiates with the activation of the macro-diversity procedure Inter-system handover: This procedure enables the handover between radio access networks that support different radio access technologies, e.g. between a UTRAN and a GSM BSS. SRNS relocation: This procedure is typically executed after an inter-RNS soft handover. In such cases, the role of the serving RNS needs to be transferred to another RNS in order to avoid the inefficient resources utilisation within the UTRAN. The SRNS relocation implies that the Iu interface connection point is relocated from one RNS to another.

3.4.2.4 Radio resource management The radio resource management is concerned with the allocation and maintenance of radio communication resources. The allocation procedure encompasses: † the selection of some communication resources from a pool of resources, and † the signalling procedures to allocate the selected resources to the mobile that requested for them.

The maintenance procedure aims to guarantee that the allocated communication resources will provide the requested quality of service. This procedure is very important in mobile communications systems where the radio environment is subject to continuous changes and to interference. In a sense, the radio resource management entity maintains a pool of communications resources and receives requests for allocation of new resources or release of already

176

Broadband Wireless Mobile: 3G and Beyond

used resources. These requests may originate from the users, who request new calls or release established calls, or from the maintenance process, which modifies the resources allocated to a user to guarantee accepTable 3.communications quality. Some functions related to radio resource management are listed below: † † † † † †

radio bearer connection setup and release; allocation and de-allocation of physical radio channels; rf power control and setting; radio channel coding control and setting. In the following two sections we identify two important aspects of the UTRAN: The protocol architecture that provides for the separation of the control and the user planes in UTRAN, and † the levels of association between the UE and the UTRAN.

3.4.3 Control and User Plane Separation in UTRAN Figure 3.11 illustrates the protocol architecture between the UE and the SRNC. This figure hides any details pertaining to Iub and Iur interfaces. However, it is implicitly assumed that these interfaces also support separate protocol stacks for the control and the user planes. On the UE side, the NAS signalling messages (i.e. signalling messages for the CN) is handled by the RRC protocol, whereas the user-data information is handled by the PDCP protocol. These protocols, along with the RLC, MAC and the L1 layer are discussed in Chapter 2. On the SRNC side, the NAS signalling messages are relayed from the RRC to the RANAP and subsequently transported to the appropriate CN domain through the Iu signalling transport layer. The user data is relayed from the PDCP layer to the GTP-U layer and subsequently to the appropriate CN domain through the Iu data transport layer. The details of the Iu signalling transport and Iu data transport, as well as the details of RANAP, will be discussed later on, in the context of the Iu protocol architecture discussion.

Figure 3.11

The separation of user plane and control plane in UTRAN.

Network Architecture

177

It is evident that the control-plane and user-plane transmissions pass through different and separate protocol stacks in each UTRAN node. In addition, in the UTRAN transport networks, which are typically based on ATM technology, there are different ATM virtual circuits for the control and user plane traffic. Usually, the virtual circuits carrying control plane information are given priority with respect to the other virtual circuits. 3.4.4 UE–UTRAN Association The level of association between a UE and the UTRAN depends on whether there is a signalling connection between the UE and the UTRAN. When such a signalling connection exists, the mobile is formally termed to be in UTRAN-CONNECTED mode. Otherwise, when no such signalling connection exists, the mobile is in UTRAN-IDLE mode. In UTRAN-IDLE mode, there is virtually no association between the UTRAN and the UE. In fact, from UTRAN’s point of view, the UE does not exist, that is, there is no RRC connection established (see Chapter 2). In this mode, the UE cannot start uplink transmission and it is not assigned any downlink radio resources. The UE merely camps on a UTRAN cell and decodes some downlink information, such as system configuration and access information. If the UE has previously registered to a CN domain, then it also decodes the broadcast paging information for identifying potential page messages for it. In general, a UE in UTRAN-IDLE mode performs the cell select and reselect procedures and all the other procedures specified in TS 25.304 [30] and in TS 23.122 [31]. The UTRAN-CONNECTED mode is entered as soon as the RRC connection is established. In such situations, the UE establishes an association with the UTRAN and it is assigned a Radio Network Temporary Identity (RNTI) to be used as UE identity on the common transport channels. In UTRAN-CONNECTED mode the UE is always in one of the four following RRC states: CELL DCH, CELL FACH, CELL PCH and URA PCH. The RRC state of the UE depends on the UE level of activity on the radio interface and on the requested quality of service. Typically, in the CELL DCH state dedicated downlink and uplink resources are allocated to UE, for instance, to support a conversational call. In the CELL FACH state, downlink and uplink activity is possible for the UE but no dedicated radio resources are allocated and, therefore, the UE uses the uplink and downlink common transport channels (see Chapter 2). These channels provide an effective means of multiplexing a large number of UEs, which are not engaged in calls with real-time requirements. In the CELL FACH state the UE monitors a downlink common transport channel and receives any protocol units that contain its RNTI in the address identifier. Typically, a UE that runs a web application could be in CELL FACH state, rather than in CELL DCH state. In CELL PCH and URA PCH states, no uplink and downlink resources are allocated to the UE and, consequently, no uplink transmission is possible. In these states, the UE decodes the broadcast information and monitors the appropriate paging channels. The difference between the CELL PCH and the URA PCH states is that in CELL PCH the position of the UE is known to the UTRAN at the cell level, whereas in URA PCH the position of the UE is known to the UTRAN at the UTRAN Registration Area (URA) level. When a new RRC connection is established and the UE enters the UTRAN-CONNECTED mode, the UE is allocated radio resources based on the type of the requested service. Typically, for conversational and streaming type of calls the UE is assigned dedicated resources and therefore is in CELL DCH state. On the other hand, for background types of calls the UE

178

Broadband Wireless Mobile: 3G and Beyond

may be assigned common uplink and downlink radio resources and therefore be in CELL FACH state. Typically, a UE in CELL FACH state may transit to CELL PCH and subsequently in URA PCH state when it is inactive for long time duration. The UE leaves the UTRAN-CONNECTED mode and returns to the UTRAN-IDLE mode as soon as the RRC connection is released. 3.4.5 The Uu Interface As already mentioned, this interface is based on the W-CDMA technology. This technology is used mainly as a radio multiplexing technology, which deals with the way the radio resources are divided and allocated to several mobile users. However, W-CDMA also specifies some aspects related to the radio transmission technology. In particular, the radio transmission technology should be based on spread spectrum techniques. This is in contrast with other multiplexing technologies, such as TDMA and FDMA, which typically do not impose any particular requirements to the radio transmission scheme. The details of the Uu interface are provided in Chapter 2. 3.4.6 The Iu Interface As we have already seen, the Iu interface is used to connect UTRAN or GERAN to CN. Strictly speaking, Iu is not an interface. This is because the term interfaceis typically used to refer to a point-to-point connection, whereas, Iu is rarely (if ever) configured as a point-topoint connection. For this reason, the interconnection between the UTRAN and the CN is referred to in the specifications as Iu reference point. Nevertheless, it is customary (and more convenient) to also use the term Iu interface, although it is not strictly correct. In this chapter, we use the terms interface and reference point interchangeably. As depicted in Figure 3.12, the Iu interface is logically divided into two interfaces: Iu-cs and Iu-ps. The Iu-cs refers to the logical connection between the RAN and the CS domain, whereas the Iu-ps refers to the logical connection between the RAN and the PS domain. It is important to note that the signalling procedures across these two logical interfaces are identical, i.e. there is a common signalling protocol, called Radio Access Network Application Part (RANAP). However, the user-plane protocols are different in Iu-cs and Iu-ps. The Iu interface has been specified so as to satisfy the following requirements: † The Iu must support all service capabilities offered to UMTS users, including:

dedicated circuits, especially for voice; best-effort packet services (e.g. Internet/IP); real-time multimedia services requiring a higher degree of QoS (CS/PS based); UMTS signalling and backward compatibility towards GSM signalling scheme. † The Iu must support transfer of transparent non-access signalling between the UE and the CN. † The Iu must support separation of each UE on the protocol level. † The Iu must support procedures to establish, maintain and release various types of Iu bearers. † The Iu must support procedures for Intersystem handover.

179

Network Architecture

Figure 3.12

Iu interface logical configuration.

In the rest of this section we will discuss the key features of the Iu interface, the protocol architecture used over the Iu interface and the signalling procedures across this interface.

3.4.7 Key Features of Iu Interface 3.4.7.1 Independence from the RAN technology In contrast to previous RAN-CN interface technologies (like the Gb interface), Iu interface is independent of the particular features of a given RAN. In this sense, the core network needs to know nothing about the structure of the RAN and the corresponding technology used. To further elaborate on this, we consider the following example. Assume, for instance, that the core network needs to send a signalling message to a particular mobile. Therefore, the core network will send a message to the RAN along with some addressing information that will identify the target mobile. Clearly, if the addressing information contains parameters such as the target site id, the target cell id, the id of the target protocol within the mobile, etc., then the core network is bound to using RAN-specific parameters and concepts. In this case, if the RAN technology is modified, corresponding modifications will be required in the core network to cope with the change of parameters and change of parameter semantics. If, on the other hand, the addressing information contains only a generic user identifier, the core network is made independent of the RAN. In this case, the same core network can be connected to several RANs, each one possibly based on different RAN technology. 3.4.7.2 Independence between the transport network and radio access network As indicated in Figure 3.13(a), the radio access network is typically connected to the core

180

Broadband Wireless Mobile: 3G and Beyond

Figure 3.13

Typical physical implementation of Iu. (b) The model of Iu functional strata.

network through a transport network. In general, the latter can be based on ATM, IP or any other technology. Having this in mind, we can logically divide the protocol architecture across Iu interface into two layers: the transport network layer or transport stratum and the radio network layer or radio stratum (see Figure 3.13(b)). This division is consistent with the aforementioned separation of UTRAN architecture into the radio network layer and the transport network layer (see section 3.4.1). The transport stratum depends on the architecture of the transport network, while the radio stratum depends on the architecture of the radio access network (e.g. UTRAN). In the control plane the radio stratum is composed by the RANAP protocol, while in the user plane, it is composed by the appropriate user plane protocols of the CS and PS domains. The functionality of the transport stratum corresponds to the functionality of the four first layers of the OSI model, i.e. physical, data link, network, and transport. This means that the transport stratum implements procedures such as sequencing, fragmentation/reassembly, error checking, data retransmission, flow control, etc. Wellknown protocols that provide such functionality include TCP and UDP. One key feature of Iu interface is that the transport and radio strata are independent to each other. The goal of this independence is to make it feasible to: † evolve transport and radio network technologies independently of each other, and † use the transport technology that suits the operator’s preferences.

181

Network Architecture

Figure 3.14

General protocol architecture of Iu logical interfaces.

3.4.7.3 Separation between control-plane traffic (i.e. signalling) and user-plane traffic Iu interface satisfies one of the fundamental principles of 3G systems: the separation between the control-plane traffic and the user-plane traffic. This means that the control-plane traffic is transported by using protocols and transport channels that are different and independent from the corresponding protocols and channels used to transport the user-plane traffic. For instance, when the transport technology is based on ATM, different virtual circuits could be used for control and user plane traffic. As we have already discussed, the separation of control and user plane is applied in the entire UMTS architecture and not only to Iu interface. By taking into account the last two headings, the generic architecture of Iu-cs and Iu-ps interfaces in the control and user plane can be illustrated as in Figure 3.14. Observe that Iu-cs and Iu-ps share a common architecture in the control plane, whereas they feature distinct architectures in the user plane. In the following section we explain in more detail these architectures.

3.4.8 Protocol Architecture across Iu As outlined in the previous section, across the Iu interface the transport stratum used to transport control-plane messages could be different from the transport stratum used to transport user-plane data. In other words, the transport technology could be different between the user-plane and the control-plane. In the following two subsections we discuss the protocol architecture of the transport stratum as used in the control and user planes. This discussion will be part of the discussion of the overall protocol architecture employed across the Iu-cs and Iu-ps interfaces.

182

Broadband Wireless Mobile: 3G and Beyond

3.4.8.1 Protocol Architecture across Iu-cs Figure 3.15 illustrates the protocol architecture employed across the Iu-cs interface. In the vertical direction, we distinguish the control plane and the user plane. These planes include the protocol stacks used to transfer network control messages (i.e. signalling) and user data respectively. In the horizontal direction, we distinguish the radio stratum and the transport stratum. The transport stratum is typically based on ATM technology and is merely used to facilitate the reliable transport of upper-layer messages (both user-plane and control-plane) between network nodes. On the other hand, in the radio stratum we find the protocols, which are specific to the RAN technology, and are used (1) to implement the Iu signalling procedures (in the control plane) and (2) to process the user data (in the user plane). As indicated in Figure 3.15, the user-plane traffic across Iu-cs interface is supported over the ATM Adaptation Layer 2 (AAL2) [42]. This means that each call towards the CS domain is supported by a specific AAL2 virtual circuit, which is set up before the call is established and is configured to match the requested QoS profile. In order to dynamically set up and release AAL2 virtual circuits, there is a need for a control plane in the transport stratum. This plane is depicted in the middle of Figure 3.15 and is referred to as transport stratum control plane. The AAL2 signalling protocol, which is used for establishing AAL2 connections towards the CS domain, is specified in ITU-T recommendation Q.2630.1 [43]. The protocols below Q.2630.1 develop a particular transport bearer, suitable for transporting the Q.2630.1 signalling messages. Q.2150.1 [44] is referred to as ‘ATM Adaptation Layer-Signalling Transport Converter for the MTP3b’ and provides the necessary adaptation functions for operating Q.2630.1 over the MTP3b layer.

Figure 3.15

Iu cs protocol architecture.

Network Architecture

183

MTP3b provides message routing, discrimination and distribution (for point-to-point link only), signalling link management load sharing and changeover/back between links within one link-set. The following three sub-layers, SSCF-NNI [39], SSCOP [38] and AAL5 [41], consist the SAAL-NNI, i.e. the signalling ATM Adaptation Layer for Network-to-Network Interface. The Service-Specific Co-ordination Function (SSCF) maps the requirements of the above layer to the requirements of SSCOP. It also provides SAAL connection management, link status and remote processor status mechanisms. Further details are provided in ITU-T’s recommendation Q.2140 [39]. The Service Specific Co-ordination Protocol (SSCOP) is a peer-to-peer protocol, which provides mechanisms for the establishment and release of connections and the reliable exchange of signalling information between signalling entities. In particular, SSCOP provides the following functions: † † † † † † † † †

transfer of messages with sequence integrity; error correction by selective retransmission; flow control; connection control; error reporting to layer management; connection maintenance in the prolonged absence of data transfer; local data retrieval by the user; error detection of protocol control information, and status reporting.

AAL5 is the ATM Adaptation Layer 5 and is specified in I.363.5 [41]. It is a simple adaptation layer that receives a block of data from the upper layer and † appends a trailer, which includes a length indicator and a CRC, † if necessary, appends some bits to make the length of the whole block plus the trailer an integral of 48 (each ATM cell carries a payload of 48 octets), and † segments the resulted AAL5 block into an integral number of ATM cells, which are sent to the lower layer for transmission.

At the receiving side, the AAL5 reassembles all the ATM cells belonging to the same upper-layer block, removes any padding information and checks the CRC to identify potential errors. Blocks that pass the CRC check are forwarded to the upper layer. As shown in Figure 3.15, the signalling transport, i.e. the protocol stack used to transport the signalling (RANAP) messages, is composed of the Signalling Connection Control Part (SCCP) [32 36], MTP3b and SAAL-NNI. SCCP is the main signalling protocol used in SS7 signalling networks. It operates on top of MTP3 or MTP3b and enhances the functions provided by the lower layers (e.g. MTP3). The key enhancements provided by SCCP are outlined below: Connectionless signalling transport services This type of service transports messages between SCCP peers with no flow control, no sequencing, and no segmentation. Therefore, the correct delivery of messages is not guaranteed. The maximum size of a message (or, a network service data unit, NSDU) can be up to

184

Broadband Wireless Mobile: 3G and Beyond

255 octets, when addressing is not by global title. Across Iu interface, the connectionless service of SCCP is used to implement the Iu general control services and the Iu notification services. The general control services are services related to the whole Iu interface instance between an RNC and the CN. For instance, the Reset procedure used to initialise the UTRAN in the event of a failure in the CN or vice versa, is one of the procedures used to implement the general control services. On the other hand, the notification services are services related to a specific mobile or all the mobiles in a specified area. A typical example is the paging procedure (see below), which is one of the procedures used to implement the Iu notification service. Connection-oriented signaling transport services With this type of service, the SCCP messages exchanged between two SCCP peers are subject to sequencing, retransmission and segmentation. In this way, a reliable and in-sequence signalling transport is established. Across Iu interface, the SCCP connection-oriented services are used to support the Iu dedicated control services, which are services related to one single mobile. Procedures used to implement the dedicated control services include the relocation procedure, the Iu release procedure, the direct transfer procedure, etc. Details on these procedures are provided in subsequent sections. Enhanced addressing capability The SCCP gives its users the option to use a global title (e.g. an E.164 number) as an address that uniquely identifies an application residing in a remote SS7 node. In such cases, SCCP performs a translation function (referred to as Global Title Translation, GTT), which translates the global title into an SS7 point code (PC) and subsystem number (SSN). The PC, SSN pair are subsequently used by MTP3 to route the message to the appropriate SS7 node. By using global titles as an addressing scheme, the SCCP users do not need to know the exact SS7 addresses (e.g. PC, SSN) of the entities they want to communicate with. In addition to the global title support, SCCP can address up to 255 users above it (by employing individuals SSNs), whereas MTP can only address up to 16 MTP users. RANAP is one of the SCCP users and it is assigned the SSN number 142. From the above discussion we conclude that the physical implementation of the Iu-cs interface could be as the one illustrated in Figure 3.16. In this figure, there are N AAL2 virtual circuits (VCs), each one used to support an individual user call towards the CS domain. In addition, there are two preconfigured AAL5 VCs: † One transport signalling VC, which typically has a virtual path identifier of 0 and a virtual channel identifier of 5. This VC is used by the transport stratum control plane, that is, it is used to transfer Q.2630.1 messages to and from the ATM network. These messages are used to dynamically manage the AAL2 VCs. † One radio signalling VC that transports signalling (i.e. RANAP) messages between the RNC and the MSC.

3.4.8.2 Protocol architecture across Iu-ps Figure 3.17 illustrates the protocol architecture used across the Iu-ps. When comparing this

185

Network Architecture

Figure 3.16

Figure 3.17

Typical ATM VCs on Iu cs interface.

Protocol architecture across Iu ps.

186

Broadband Wireless Mobile: 3G and Beyond

Figure 3.18 AAL5/ATM layers are used to establish VCs between the RNC and the PS domain. (b) The concept of GTP U tunnels.

protocol architecture with the corresponding protocol architecture of Iu-cs (see Figure 3.16), three observations can be made: The user-plane transport in Iu-ps is not based on AAL2, but on AAL5 Also, since on top of AAL5 there is IP, UDP and GTP-U, we conclude that user data across Iu-ps is routed with IP and is transported in GTP-U tunnels. Figure 3.18 explains schematically how data is transferred over the Iu-ps. As shown in Figure 3.18(a), the AAL5/ATM layers are used to establish a virtual channel between the RNC and the PS domain through the ATM network. This virtual circuit is used to multiplex all IP traffic. In the PS domain the AAL5 VC typically terminates to an IP router, which routes the IP packets to and from the correct SGSN in the PS backbone. As shown in Figure 3.18(b), the IP traffic between the RNC and the SGSN is logically decomposed into a number of individual GTP-U tunnels. These tunnels are managed by the RANAP protocol, which contains appropriate procedures to setup new tunnels, or release already established tunnels. In particular, a new GTP-U tunnel is established when a new RAB is established (see section 0 for details). The GTP-U tunnel each packet belongs to is identified by the tunnel endpoint identifier that is included in the GTP-U header. One mobile may utilise one or more GTP-U tunnels, each one corresponding to a different set of protocol parameters (e.g. IP address) and/or QoS parameters.

Network Architecture

187

In Iu-ps there no transport stratum control plane. Indeed, each Iu bearer (i.e. each GTP-U tunnel) across Iu-ps does not require a specific ATM virtual circuit to be established. As shown in Figures 3.18(a) and (b), all GTP-U tunnels across Iu-ps are multiplexed into one (or more) AAL5 VCs, which have been preconfigured by a management entity. Therefore, the bearer traffic over Iu-ps uses already established ATM VCs and there is no need to dynamically establish and configure new ones. There are two options for the transport of signalling traffic over Iu-ps (Figure 3.17) The first one is identical to the signalling transport employed on the Iu-cs and is composed of the protocols SCCP/MTP3b/SSCF-NNI/SSCOP. The second one is based on IP-over-ATM, rather than on MTP-over-ATM, and is composed of the protocols SCCP/M3UA/SCTP/IP. M3UA is the MTP3 User Adaptation layer specified by IETF [45]. It is used to maps the services required by SCCP to the services provided by SCTP [46]. Note that SCCP assumes it runs on top of MTP3 (or MTP3b). M3UA provides the interworking functionality needed to have SCCP run on top of SCTP. SCTP is the Stream Control Transport Protocol, which is also specified by IETF, currently in a draft Internet specification [46]. SCTP is a reliable transport protocol operating on top of a connectionless packet network such as IP. It offers the following services: † † † † †

Error-free non-duplicated transfer of messages; Data fragmentation and reassembly; Sequenced delivery; Optional bundling of multiple messages into a single SCTP packet, and Network-level fault tolerance through supporting of multi-homing at either or both ends of an association.

Observe that SCTP functionality is similar to the functionality of SSCOP. A physical implementation of the Iu-ps interface could be as the one illustrated in Figure 3.19. In this figure, there is one radio signalling VC that transports signalling (i.e. RANAP) messages between the RNC and the SGSN, and one VC (or more) used as bearer transport, i.e. used to transport user packets between the RNC and the SGSN. As discussed above, the context associated with each user packet is identified by the GTP-U tunnel used to transfer this packet.

3.4.9 Signalling Procedures across Iu Since the signalling protocol across Iu interface is RANAP, all signalling procedures across the Iu interface are effectively performed by means of RANAP procedures. Therefore, the signalling procedures discussed in this section are RANAP procedures and are described in more detail in 3GPP TS 25.413 [22]. RANAP supports a number of functions including the following: † Management of radio access bearers (RABs), i.e. setting up, modifying and releasing RABs between a mobile and the core network. In general, the RAB management is performed by the core network, however, the RAN has the capability to request the release of RABs.

188

Broadband Wireless Mobile: 3G and Beyond

Figure 3.19

Typical ATM VCs across the Iu ps interface.

† Managing of the overall Iu resources and in particular:

Explicitly release of all resources related to one Iu connection; the Iu release is typically managed by the CN, however, the RAN has also the capability to request the release of all Iu connection resources from the corresponding Iu connection; Controlling the load of the Iu interface; Resetting the Iu interface. † Mobility management functions, such as:

Relocating the serving RNC, that is, moving the Iu resources from one RNC to another; SRNS context forwarding, that is, transferring the SRNS context from the RNC to the CN for intersystem forward handover. † Paging a given mobile. † Transport of non-access stratum signalling messages between the mobile and the CN. † Managing the security mode in the RAN, such as sending the security keys and setting the operation mode for security functions. † Reporting the data volume (for specific RABs) which has been unsuccessfully transmitted from the RAN to the mobile. † Tracing the activity of a given mobile. † Reporting the location of a given mobile. † Reporting of general error situations.

Table 3.1 lists all the RANAP functions and the RANAP elementary procedures utilised in the context of each function. The RANAP functions outlined above are implemented by means of one or more RANAP

189

Network Architecture Table 3.1

RANAP elementary procedures categorised according to their functional context

RANAP function Management of RABs Management of Iu Resources SRNS Relocation

SRNS Context Forwarding

Paging Transport of non access stratum signalling messages between the mobile and the CN Management of security mode Reporting of data volume Tracing the activity of a given mobile Reporting the location of a given mobile Miscellaneous

RANAP elementary procedures RAB Assignment Request RAB Assignment Response RAB Release Request Iu Release Request Iu Release Command Iu Release Complete Relocation Required Relocation Request Relocation Request Acknowledge Relocation Command Relocation Detect Relocation Complete Relocation Preparation Failure Relocation Failure Relocation Cancel Relocation Cancel Acknowledge SRNS Context Request SRNS Context Response SRNS Data Forward Command Forward SRNS Context Paging Direct Transfer Security Mode Command Security Mode Complete Security Mode Reject Data Volume Report Request Data Volume Report CN Invoke Trace CN Deactivate Trace Location Reporting Control Location Report Common ID Overload Error Indication Reset Reset Acknowledge Reset Resource Reset Resource Acknowledge

elementary procedures. Table 3.2 lists all RANAP elementary procedures and the RANAP messages associated with each elementary procedure. In the following sections we discuss some examples that illustrate how some common RANAP functions are provided. 3.4.9.1 Transport of NAS signalling messages between the mobile and the CN As long as a mobile communicates with the core network, there is a unique logical connection across the Iu interface devoted to the transmission of signalling messages between that mobile and the core network. This logical connection is referred to as Iu signalling connection. Since this applies to all the mobiles that have attached to the CN, the Iu signalling circuit between the RAN and the CN features a number of Iu signalling connections, each one devoted to one and only one mobile. This is illustrated in Figure 3.20. As explained below, each Iu signalling connection effectively corresponds to a SCCP connection. One of the functions provided by the Iu interface is to establish an Iu signalling connection per mobile. Let us explain how this is accomplished. Once the mobile has established a signalling connection with the RAN, it may send the fist NAS message to CN. This message

IU RELEASE COMMAND RELOCATION REQUIRED

RELOCATION REQUEST

Iu Release Relocation Preparation

Relocation Resource Allocation Relocation Cancel

RAB RELEASE REQUEST IU RELEASE REQUEST

RAB Release Request Iu Release Request

Reset Reset Resource

Data Volume Report

Security Mode Control

SRNS CONTEXT REQUEST SECURITY MODE COMMAND DATA VOLUME REPORT REQUEST RESET RESET RESOURCE

SRNS Context Transfer

RELOCATION CANCEL

Initiating message

RANAP elementary procedures and messages

Elementary procedure

Table 3.2

RESET ACKNOWLEDGE RESET RESOURCE ACKNOWLEDGE

IU RELEASE COMPLETE RELOCATION COMMAND RELOCATION REQUEST ACKNOWLEDGE RELOCATION CANCEL ACKNOWLEDGE SRNS CONTEXT RESPONSE SECURITY MODE COMPLETE DATA VOLUME REPORT

Response message on successful outcome

SECURITY MODE REJECT

RELOCATION PREPARATION FAILURE RELOCATION FAILURE

Response message on unsuccessful outcome

190 Broadband Wireless Mobile: 3G and Beyond

Location Report Initial UE Message Direct Transfer Overload Control Error Indication

SRNS Data Forwarding Initiation SRNS Context Forwarding from Source RNC to CN SRNS Context Forwarding to Target RNC from CN Paging Common ID CN Invoke Trace CN Deactivate Trace Location Reporting Control

Relocation Detect Relocation Complete

RELOCATION DETECT RELOCATION COMPLETE SRNS DATA FORWARD COMMAND FORWARD SRNS CONTEXT FORWARD SRNS CONTEXT PAGING COMMON ID CN INVOKE TRACE CN DEACTIVATE TRACE LOCATION REPORTING CONTROL LOCATION REPORT INITIAL UE MESSAGE DIRECT TRANSFER OVERLOAD ERROR INDICATION

Network Architecture 191

192

Figure 3.20 CN.

Broadband Wireless Mobile: 3G and Beyond

The Iu signaling channel is used to carry MS specific signaling between the mobiles and

could for instance be a request to attach to the core network, or a request to update the location/routing area. When the RAN receives this first NAS message it identifies that currently there is no Iu signalling connection for that specific mobile. Therefore it establishes a new Iu signalling connection across the Iu interface and binds this connection to that mobile. This new Iu signalling connection is created by the SCCP layer and therefore it corresponds to a new SCCP connection. From the above discussion it is clear that the first NAS message sent by the mobile triggers the establishment of a new Iu signalling connection across the Iu interface. This new connection will carry all subsequent signalling messages exchanged between the mobile and CN. This procedure is illustrated in Figure 3.21. Note that the first NAS message is transferred within an Initial UE Message, while every subsequent NAS message, either from or to the CN, is transferred within a Direct Transfer message. 3.4.9.2 RAB setup/release Assume now that after successfully attaching to the core network, the mobile needs to request a service, e.g. to establish a voice call via the CS domain. To support this new call, a new RAB needs to be established across Iu-cs. In our discussion here, it is not important when exactly this new RAB is established, i.e., if it is established before or after the call is answered. It is important however to understand what signalling procedures are executed on Iu-cs interface to establish a RAB suitable to support the requested call. To establish a new RAB, or to modify/release an existing RAB, the RAB Assignment procedure is executed on Iu interface (this procedure applies to both Iu-cs and Iu-ps). This procedure is schematically illustrated in Figure 3.22 and it is explained below. In our example, what triggers the RAB Assignment procedure is the request received by the MSC to establish a new voice call. The type of the call requested is described in the request sent by the mobile and it is used by the network to reserve and configure the appropriate transport resources. When the MSC receives the mobile’s request, it sends a RAB Assign-

193

Network Architecture

Figure 3.21 Establishment of a new Iu signaling connection.

Figure 3.22 Signaling for RAB establishment.

194

Broadband Wireless Mobile: 3G and Beyond

ment Request message to the RNC. This message (1) commands the RNC to establish the corresponding radio bearer between the RNC and the mobile and (2) identifies the information required by the RNC to setup a new AAL2 channel, which is going to support the requested RAB. Among others, this information includes a QoS profile, a transport addressand a transport identifier. In response, the RNC signals the ATM network to establish a new AAL2 channel with the appropriate QoS characteristics, e.g. capacity, error rate, etc. The transport address is used to identify the ATM address of the target endpoint (the MSC in this case). The transport identifier is used at the RNC to associate the new AAL2 channel with the requested RAB. That is, any user data that should be transported on this RAB is sent to the transport channel identified by the transport identifier. Subsequently, the RNC establishes the appropriate radio bearer (i.e. one that satisfies the requested QoS profile) and returns a RAB Assignment Response message to the MSC. If the RNC cannot establish an appropriate radio bearer, it sends a RAB Assignment Response message with a cause indicating why the radio bearer couldn’t be established, e.g. indicating that the requested maximum bit rate is not available. In this case, the MSC may re-attempt another RAB Assignment Request message with a different QoS profile. When the RNC cannot satisfy the requested RAB assignment, it may queue the request and report the outcome of the establishment later. When a RAB is to be established between a mobile and the PS domain, the above procedure is similar. In this case, however, there is no need to dynamically set up a new AAL2 virtual channel. For this reason, the transport address and transport identifier have a different meaning. In the RAB Assignment Request the transport address corresponds to the IP address of the SGSN and the transport identifiercorresponds to a GTP tunnel identifier at SGSN’s side. On the other hand, in the RAB Assignment Response the transport address corresponds to the IP address of the RNC and the transport identifiercorresponds to a GTP tunnel identifier at RNC’s side. This way, after the RAB assignment a new GTP tunnel is established to transfer the user data. 3.4.9.3 Relocation procedure The Relocation procedure is executed when the Iu interface needs to be relocated from one RNC to another and possibly from one CN element to another. The relocation is needed to support the roaming of a mobile between geographical areas controlled by different RNCs, and possibly by different SGSNs and/or MSCs. When the mobile enters an area, which is controlled by a different RNC but by the same SGSN (or MSC), the relocation is referred to as intra-SGSN relocation (or inta-MSC relocation). On the other hand, when the mobile enters an area, which is controlled by a different RNC and a different SGSN (or MSC), the relocation is referred to as inter-SGSN relocation (or inter-MSC relocation). In every relocation case, the role of the serving RNC is transferred from one RNC (the source RNC) to another (the target RNC). For this reason, the relocation is also called (intra-/inter-SGSN/MSC) serving RNS relocation. Before relocation, the source RNS functions as a serving RNS, while after relocation, the target RNS becomes the serving RNS. Figure 3.23 illustrates the situation before, during and after the relocation. Figure 3.23(a) shows the situation before the need for relocation. In this case, the RNS1 functions as the serving RNS. In Figure 3.23(b) the mobile has moves into the area covered by the RNS2.

Network Architecture

195

Figure 3.23 Soft handover: (a) Traffic flow before the handover and (b) Traffic flow after the hand over.

196

Broadband Wireless Mobile: 3G and Beyond

Figure 3.23(c)

Traffic flow after the SRNS relocation.

RNS1 still connects the data path to the core network and therefore still serves as the serving RNS. On the other hand, the radio interface is handled by RNS2, which, in this case, serves as a drifting RNS for the mobile. In such a situation, a need arises to relocate the serving RNS from RNS1 to RNS2. The goal is to optimise the utilisation of the network resources and to eliminate the inefficient routing of user data over the Iur interface. For this purpose, the RNS1 will initiate the serving RNS Relocation, which after completion will lead to the situation depicted in Figure 3.23(c). In this figure, the RNS2 has become the serving RNS and connects the data path to the core network. Also, the Iur interface is no longer needed to support the transportation of user data. Figure 3.24 illustrates the message sequence diagram that corresponds to an intra-SGSN or intra-MSC relocation procedure. As it is shown, it is composed of four phases: Relocation Preparation (P1), Relocation Resource Allocation (P2), Relocation Detect (P3) and Relocation Complete (P4). The Relocation Preparation beginnings when the source RNC sends a RELOCATION REQUIRED command to the CN. This command includes: † A unique identifier to identify the source RNC and another unique identifier to identify the target RNC. † An appropriate cause value for the Relocation, to indicate e.g. whether the relocation is required for resources optimisation, for radio reasons, etc. † An indication to specify whether the required relocation shall be executed with or without the involvement of the mobile.

197

Network Architecture

Figure 3.24

Message sequence diagram for SRNS Relocation.

† A set of parameters that shall transparently be transferred from the CN to the target RNC.

When the CN receives the RELOCATION REQUIRED command, it will try to allocate the appropriate Iu resources between the target RNC and the CN. This is when P2 begins. The CN sends the RELOCATION REQUEST command to the target RNC, which contains the information required to build the same RAB configuration as the one existing for the mobile before the relocation. When the appropriate Iu resources are allocated, the target RNC responds with a RELOCATION REQUEST ACKNOWLEDGE message and phase 2 terminates. Now, the CN needs to inform the source RNC that the preparation of the relocation is over and therefore the appropriate actions may take place to perform the actual relocation of the serving RNC. For this purpose, the CN sends the RELOCATION COMMAND to the source RNC and phase 1 terminates too. At this moment, the source RNC requests from the target RNC to proceed with the relocation by sending a RELOCATION COMMIT message. The purpose of this message is to transfer the serving RNS contexts from the source RNC to the target RNC. These contexts are sent for each concerned RAB and mainly contain the appropriate sequence numbers of the user-plane messages to be subsequently transmitted in the uplink and downlink directions. It is important to note that, before sending the RELOCATION COMMIT, uplink and downlink data transfer at the source RNC is suspended for all RABs that require loss-less relocation. After having sent the RELOCATION COMMIT message, the source RNC begins the forwarding of data for each concerned RAB. Subsequently, the target RNC sends a RELOCATION DETECT message to indicate to the CN that the execution of the serving RNS relocation has been detected (this accounts for P3). At this moment, the target RNC allocates

198

Broadband Wireless Mobile: 3G and Beyond

a new Radio Network Temporary Identifier (RNTI) to the mobile and forwards this value to the mobile with RNTI REALLOCATION message. When the mobile acknowledges the correct reception of the RNTI, the target RNC considers that the relocation procedure has been completed and responds to the CN with a RELOCATION COMPLETE message (this accounts for P4). Finally, the CN will release the Iu resources between the source RNC and the CN and this completes the relocation procedure. 3.4.9.4 Paging When the CN needs to establish communication with a specific mobile that is in idle mode, the paging procedure is used. To initiate the paging, the CN sends a PAGING message to the RNC. This message contains all the necessary information for RNC to be able to page the mobile. In particular, this message indicates: † The domain indicator, which identifies the CN domain that originates the paging request. † The NAS identifier of the mobile, i.e. the IMSI of the mobile. This identifier is used by the RNC to check whether there is already a signalling connection between that mobile and the other CN domain (the one not originating the paging request). If this is true, then the already existing signalling connection is used to transfer the paging request towards the mobile. Otherwise, the paging request is transferred, as usually, over the paging broadcast channel. † The temporary identifier of the mobile, i.e. the TMSI. This identifier is used to address the particular mobile in the paging request transmitted over the radio interface. If TMSI is not available, then IMSI is used instead. † The Paging Area identifier, which identifies the area in which the radio interface paging message shall be broadcast. This does not apply to the case where there is already signalling connection between the mobile and the other CN domain. If the Paging Area identifier is not included in the PAGING message, then the whole RNC area is used as a Paging Area. † The reason why the CN initiates the paging request. This reason is transferred transparently to the mobile.

It should be noted that each PAGING message on the Iu interface pertains to only one mobile. However, the RNC may pack several pages into one relevant paging message transmitted over the radio interface.

3.4.10 Iur Interface As we have already seen, the logical interface between two RNCs in the UTRAN is referred to as Iur. This interface can be considered as a logical point-to-point interface between two RNCs, which can dynamically be established even when there’s no physical direct connection between the two RNCs. The Iur interface supports the exchange of signalling information between two RNCs and, in addition, it may support one or more data streams, each one carrying user-plane information. In this section we discuss the key features of the Iur interface, as well as the Iur protocol architecture and the Iur signalling procedures.

199

Network Architecture

3.4.10.1 Capabilities of Iur Evidently, all the key features of the UTRAN architecture are applicable to Iur too. For instance, the separation of user and control planes and the split of protocol architecture into the Radio-Network layer and the Transport-Network layer, are also applicable in Iur. In general, the Iur interface supports the: † Mobility of a UE’s radio interface between different RNSs. To support such mobility, the Iur provides support for soft and hard handover, radio resource management and synchronisation between RNSs. † Relay of uplink dedicated transport channels (DCHs) from Iub to Iur and relay of downlink dedicated transport channels from Iur to Iub. This situation arises when a mobile is served via a DRNC, as illustrated in Figure 3.12(b). † Transport of uplink and downlink common transport channels (such as, RACH, FACH, CPCH, etc.) between the DRNC and the SRNC.

3.4.10.2 Protocol architecture across Iur The protocol architecture across Iur is depicted in Figure 3.25. By comparing this figure with Figure 3.20, we conclude that the protocol architecture of the Iur interface is identical to the protocol architecture of the Iu-cs interface. The only difference is that in Iur there are two alternative protocol stacks defined for transporting Radio-Network signalling and TransportNetwork signalling. The first protocol stack is SS7-based and it is composed of MTP3-B/ SSCF-NNI/SSCOP, while the second protocol stack is IP-based and it is composed of M3UA/ SCTP/IP. Typically, both these protocol stacks run on top of ATM. A brief discussion of these protocols is provided in the section of Iu interface.

Figure 3.25

Protocol architecture across Iur.

200

Broadband Wireless Mobile: 3G and Beyond

From the Iur protocol architecture shown in Figure 3.25 we can envision that a connection between two RNCs could typically be composed by one AAL5 virtual channel (VC), which is used to carry all Radio-Network signalling messages, and by several AAL2 VCs, which carry the user plane traffic, or the so-called data streams. This is schematically illustrated in Figure 3.26. The use of SCCP in the Radio-Network signalling transport makes it possible (1) to provide enhanced capabilities for addressing RNCs and (2) to provide connection-oriented and connectionless services to the Iur Radio-Network signalling protocol, i.e. to the Radio Network Subsystem Application Part (RNSAP). As illustrated in Figure 3.27, each RNC-RNC connection features one (or maybe more) AAL5 virtual circuit connection, which is used to exchange RNSAP messages between the two RNCs. In this virtual connection, several SCCP connections can be multiplexed. These connections are referred to as Iur signaling connections and each one carries signalling traffic pertaining to a particular UE. In other words, each RNSAP message transferred over a given SCCP connection is associated with the UE that corresponds to that SCCP connection. Therefore, RNSAP messages transferred in connected mode do not need to carry a UE identifier but rather an SCCP connection identifier. The use of a Transport-Network signalling protocol, i.e. Q.2630.1 [43], is mandatory for managing and establishing new AAL2 connections between pairs of RNCs. When an SRNC needs to switch some dedicated/common data streams to a DRNC, a series of new AAL2 virtual circuits needs to be established to support these data streams. All these virtual circuits

Figure 3.26

Typical ATM VCs across Iur interface.

201

Network Architecture

Figure 3.27

Many SCCP connections are supported by RNSAP.

are established by means of Q.2630.1 signalling. Further information about Q.2630.1 can be found in the corresponding ITU-T recommendation [43]. 3.4.10.3 Signaling procedures across Iur The signalling procedures across Iur interface are equivalent to the RNSAP procedures. The execution of the RNSAP procedures provides for the following functions [23]: † Radio Link Management: This function allows the SRNC to manage radio links using dedicated resources in a DRNS. † Physical Channel Reconfiguration: This function allows the DRNC to reallocate the physical channel resources for a radio link. † Radio Link Supervision: This function allows the DRNC to report failures and restorations of a Radio Link. † Compressed Mode Control [FDD]: This function allows the SRNC to control the usage of compressed mode within a DRNS. † Measurements on Dedicated Resources: This function allows the SRNC to initiate measurements on dedicated resources in the DRNS. The function also allows the DRNC to report the result of the measurements. † DL Power Drifting Correction [FDD]: This function allows the SRNC to adjust the DL power level of one or more radio links in order to avoid DL power drifting between the Radio Links. † CCCH Signaling Transfer: This function allows the SRNC and DRNC to pass information between the UE and the SRNC on a common control channel (CCCH) controlled by the DRNS. † Paging: This function allows the SRNC to page a UE in a URA or a cell controlled by the DRNS. † Common Transport Channel Resources Management: This function allows the SRNC to utilise Common Transport Channel Resources within the DRNS (excluding DSCH resources for FDD). † Relocation Execution: This function allows the SRNC to finalize a Relocation previously prepared via other interfaces (see the Iu interface section). † Reporting of General Error Situations: This function allows reporting of general error situations, for which function specific error messages have not been defined.

202

Broadband Wireless Mobile: 3G and Beyond

† DL Power Timeslot Correction [TDD]: This function enables the DRNS to apply an individual offset to the transmission power in each timeslot according to the downlink interference level at the UE.

Table 3.3 lists all RNSAP functions and the corresponding RNSAP elementary procedures utilised in the context of each function. Table 3.4 lists all the RNSAP elementary procedures and the RNSAP messages associated with each elementary procedure. Table 3.3 RNSAP functions and corresponding procedures RNSAP function

RNSAP elementary procedures

Radio Link Management

X Radio Link Setup X Radio Link Addition X Radio Link Deletion X Unsynchronised Radio Link Reconfiguration X Synchronised Radio Link Reconfiguration Preparation X Synchronised Radio Link Reconfiguration Commit X Synchronised Radio Link Reconfiguration Cancellation X Radio Link Pre emption X Physical Channel Reconfiguration X Radio Link Failure X Radio Link Restoration X Radio Link Setup X Radio Link Addition X Compressed Mode Command X Unsynchronised Radio Link Reconfiguration X Synchronised Radio Link Reconfiguration Preparation X Synchronised Radio Link Reconfiguration Commit X Synchronised Radio Link Reconfiguration Cancellation X Dedicated Measurement Initiation

Physical Channel Reconfiguration Radio Link Supervision Compressed Mode Control [FDD]

Measurements on Dedicated Resources

DL Power Drifting Correction [FDD] CCCH Signalling Transfer Paging Common Transport Channel Resources Management

X Dedicated Measurement Reporting X Dedicated Measurement Termination X Dedicated Measurement Failure X Downlink Power Control X Uplink Signalling Transfer X Downlink Signalling Transfer X Paging X Common Transport Channel Resources Initiation

X Common Transport Channel Resources Release Relocation Execution X Relocation Commit Reporting of General Error Situations X Error Indication DL Power Timeslot Correction [TDD] X Downlink Power Timeslot Control

RADIO LINK SETUP REQUEST RADIO LINK ADDITION REQUEST RADIO LINK DELETION REQUEST RADIO LINK RECONFIGURATION PREPARE RADIO LINK RECONFIGURATION REQUEST PHYSICAL CHANNEL RECONFIGURATION REQUEST DEDICATED MEASUREMENT INITIATION REQUEST COMMON TRANSPORT CHANNEL RESOURCES REQUEST UPLINK SIGNALLING TRANSFER INDICATION DOWNLINK SIGNALLING TRANSFER REQUEST RELOCATION COMMIT PAGING REQUEST RADIO LINK RECONFIGURATION COMMIT

Radio Link Setup Radio Link Addition

Downlink Signalling Transfer Relocation Commit Paging Synchronised Radio Link Reconfiguration Commit

Uplink Signalling Transfer

Synchronised Radio Link Reconfiguration Preparation Unsynchronised Radio Link Reconfiguration Physical Channel Reconfiguration Dedicated Measurement Initiation Common Transport Channel Resources Initialization

Radio Link Deletion

Initiating message

RNSAP elementary procedures and messages

RNSAP elementary procedure

Table 3.4

– – – –

– – –

RADIO LINK RECONFIGURATION FAILURE RADIO LINK RECONFIGURATION FAILURE PHYSICAL CHANNEL RECONFIGURATION FAILURE DEDICATED MEASUREMENT INITIATION FAILURE COMMON TRANSPORT CHANNEL RESOURCES FAILURE –

RADIO LINK SETUP FAILURE RADIO LINK ADDITION FAILURE

Response message on unsuccessfuloutcome

–

RADIO LINK SETUP RESPONSE RADIO LINK ADDITION RESPONSE RADIO LINK DELETION RESPONSE RADIO LINK RECONFIGURATION READY RADIO LINK RECONFIGURATION RESPONSE PHYSICAL CHANNEL RECONFIGURATION COMMAND DEDICATED MEASUREMENT INITIATION RESPONSE COMMON TRANSPORT CHANNEL RESOURCES RESPONSE –

Response message on successful outcome

Network Architecture 203

Error Indication Downlink Power Timeslot Control [TDD] Radio Link Pre-emption

Dedicated Measurement Reporting Dedicated Measurement Termination Dedicated Measurement Failure Downlink Power Control [FDD] Compressed Mode Command [FDD] Common Transport Channel Resources Release

COMPRESSED MODE COMMAND COMMON TRANSPORT CHANNEL RESOURCES RELEASE REQUEST ERROR INDICATION DL POWER TIMESLOT CONTROL REQUEST RADIO LINK PREEMPTION REQUIRED INDICATION

– – –

– – –

–

–

–

–

–

–

–

–

–

–

–

–

–

–

Response message on unsuccessfuloutcome

–

–

–

RADIO LINK FAILURE INDICATION RADIO LINK RESTORE INDICATION DEDICATED MEASUREMENT REPORT DEDICATED MEASUREMENT TERMINATION REQUEST DEDICATED MEASUREMENT FAILURE INDICATION DL POWER CONTROL REQUEST

Radio Link Restoration

–

RADIO LINK RECONFIGURATION CANCEL

Synchronised Radio Link Reconfiguration Cancellation Radio Link Failure

Response message on successful outcome

Initiating message

RNSAP elementary procedure

Table 3.4 (continued)

204 Broadband Wireless Mobile: 3G and Beyond

Network Architecture

205

3.4.11 Iub Interface The logical interface that connects a Node-B with an RNC is termed Iub. As explained below, this interface provides two main functions: † a signalling function between the Node-B and its controlling RNC, which is basically used to control the radio resources of the Node-B, e.g. allocating, releasing, or monitoring the radio channels; † a transport or relay function, which is used to relay the information carried on the various radio transport channels to/from the RNC.

3.4.11.1 Capabilities of Iub In more detail, the Iub interface provides for the following capabilities. Radio-related signalling Through the Iub interface, the RNC can activate, deactivate or monitor the status/quality of specific transport channels on the radio interface. Each time the RNC commands the activation of a radio transport channel via Iub, it identifies the particular characteristic of this channel. In addition, the RNC can use the radio-related signalling of the Iub interface to control other radio-related issues, such as to identify the information to be transmitted on the broadcast channel, or to specify the format characteristics of this channel. Relay of radio transport channels to/from the RNC One of the main functions of the Iub interface is to relay the information carried on the various radio transport channels to and from the RNC. Each transport channel on the radio interface is mapped to a so-called data stream on the Iub interface. An Iub data stream is carried on one Iub data bearer, which is typically implemented with an AAL2 VCC. The main types of data streams used on the Iub interface are the following (see Figure 3.28): † DCH data stream: A bi-directional data stream used to transport the DCH transport frames between Node-B and RNC. One Iub DCH data stream is carried on one data bearer and, therefore, for each DCH data stream a data bearer must be established over Iub. Evidently, this calls for a transport-network signalling protocol across Iub, capable of dynamically setting up and releasing Iub data bearers. As discussed in the next section, this signalling protocol is again the Q.2630.1 [43]. † RACH data stream: An uplink data stream used to transport the RACH transport frames from Node-B to RNC. One RACH data stream is carried on one Iub data bearer. † PCH data stream: A downlink data stream used to transport the PCH transport frames from RNC to Node-B. One PCH data stream is carried on one Iub data bearer. † FDD CPCH data stream: An uplink data stream used to transport the FDD CPCH transport frames from Node-B to RNC. One FDD CPCH data stream is carried on one Iub data bearer. † TDD USCH data stream: An uplink data stream used to transport the TDD USCH transport frames from Node B to RNC. A UE may have multiple USCH data streams. One TDD USCH data stream is carried on one Iub data bearer.

206

Broadband Wireless Mobile: 3G and Beyond

Figure 3.28

Main data streams on Iub and their relation with Uu transport channels.

† FACH data stream: A downlink data stream used to transport the FACH transport frames from RNC to Node-B. One FACH data stream is carried on one Iub data bearer. † DSCH data stream: A downlink data stream used to transport the DSCH transport frames from RNC to Node-B. One DSCH data stream is carried on one Iub data bearer.

We discuss below all functions supported by the Iub interface. 3.4.11.2 Protocol architecture across Iub The protocol architecture across Iub is depicted in Figure 3.29. By comparing this figure with Figure 3.20, we conclude that the protocol architecture of the Iub interface is quite similar to the protocol architecture of the Iu-cs interface. However, a major difference exists: in the Radio-Network signalling transport and in the Transport-Network signalling transport, there is neither SCCP layer nor MTP3b layer. This means that the enhanced capabilities provided by SCCP in the Iu and Iur interfaces are not needed on Iub interface. For instance, there is no need to provide connection-oriented services to NBAP, nor enhanced addressing capabilities. As we observe, NBAP signalling is carried over the Signalling ATM Adaptation layer for user-to-network interface, SAAL-UNI [37], which is composed by SSCF-UNI, SSCOP [38] and AAL5 [41] protocols. SSCF-UNI provides mainly an adaptation function, mapping the services expected by NBAP to the services provided by SSCOP. The latter is a peer-to-peer protocol, which provides mechanisms for the reliable and in-sequence exchange of NBAP messages between the Node-B and the RNC. As with Iu-cs and Iur, the Q.2630.1 is used as a Transport-Network signalling protocol, which is used for dynamic establishment and management of Iub data bearers. In the user plane of the Radio-Network layer, there are several Frame Protocols (FPs), one per radio transport channel (see [25 27] for more details on FPs). The frame protocol used for DCH channels is termed DCH FP, the frame protocol for the RACH channel is termed RACH FP, etc. In the uplink direction, a frame protocol in the

207

Network Architecture

Figure 3.29

Protocol architecture across Iub.

Node-B adds a header to every transport frame received from the radio interface and forms a FP PDU that is transported to the RNC over a suitable Iub data stream (or AAL2 connection). For instance, the RACH FP at the Node-B adds a header component into every RACH transport frame received from the RACH transport channel and forms a RACH FP PDU that is transported to the RNC over the Iub RACH data stream. Similarly, the DCH FP adds a header component into every DCH transport frame received from a DCH transport channel and forms a DCH FP PDU that is transported to the RNC over an Iub DCH data stream. In the downlink direction, a frame protocol removes the FP header and relays the payload to a radio transport channel. Each FP header contains a Frame Type field and information related to the frame type. In addition, DCH FP headers contain a CRC checksum. Typically, every FP supports two types of FP frames: the FP Data Frame and the FP Control Frame. The type of every FP frame (or PDU) is indicated by the Frame Type field. FP Data frames are used to carry transport frames coming from or going to the radio interface, whereas FP Control frames are used to implement an important in-band signalling mechanism, which is used for various purposes. For instance, the DCH FP, besides transporting DCH transport frames, provides the following additional services: † Transport of outer loop power control information between the SRNC and the Node B, † Support of transport channel synchronisation (used to achieve or restore the synchronisation of the DCH data stream in the downlink direction, and as a keep alive procedure in order to maintain activity on the Iub/Iur data bearer), † Support of Node synchronisation (used by the SRNC to acquire information on the NodeB timing),

208

Broadband Wireless Mobile: 3G and Beyond

Figure 3.30 Typical ATM VCs across Iub interface.

† Transfer of radio interface parameters from the SRNC to the Node B (used to update radio interface parameters which are applicable to all radio link of a given UE), etc.

The FP Control frames are also used for other purposes, such as: (1) to support flow control (e.g. the FACH FP supports flow control frames, used by the DRNC to control the user data flow); (2) to request and allocate capacity on demand, etc. A more thorough discussion of the frame protocols can be found in [25] and [26]. From the Iub protocol architecture shown in Figure 3.29 we can envision that a connection between a Node-B and an RNC can typically be composed by: (1) one AAL5 virtual channel (VC), which is used to carry all the Radio-Network signalling (i.e. NBAP) messages; (2) one AAL5 VC, which is used to carry all the Transport-Network signalling (i.e. Q.2630.1) messages; (3) several AAL2 VCs, which carry the Iub data streams. This is schematically illustrated in Figure 3.30. 3.4.11.3 Signaling procedure across Iub The signalling procedures across Iub interface are equivalent to the NBAP procedures. The execution of the NBAP procedures provides for the following functions [24]: † Cell Configuration Management: This function gives the CRNC the possibility to manage the cell configuration information in a Node B.

Network Architecture

209

† Common Transport Channel Management: This function gives the CRNC the possibility to manage the configuration of Common Transport Channels in a Node B. † System Information Management: This function gives the CRNC the ability to manage the scheduling of System Information to be broadcast in a cell. † Resource Event Management: This function gives the Node B the ability to inform the CRNC about the status of Node B resources. † Configuration Alignment: This function gives the CRNC and the Node B the possibility to verify and enforce that both nodes have the same information on the configuration of the radio resources. † Measurements on Common Resources: This function allows the Node B to initiate measurements in the Node B. The function also allows the Node B to report the result of the measurements. † Radio Link Management: This function allows the CRNC to manage radio links using dedicated resources in a Node B. † Radio Link Supervision: This function allows the CRNC to report failures and restorations of a Radio Link. † Compressed Mode Control [FDD]: This function allows the CRNC to control the usage of compressed mode in a Node B. † Measurements on Dedicated Resources: This function allows the CRNC to initiate measurements in the Node B. The function also allows the Node B to report the result of the measurements. † DL Power Drifting Correction [FDD]: This function allows the CRNC to adjust the DL power level of one or more Radio Links in order to avoid DL power drifting between the Radio Links. † Reporting of General Error Situations: This function allows reporting of general error situations, for which function specific error messages have not been defined. † Physical Shared Channel Management [TDD]: This function allows the CRNC to manage physical resources in the Node B belonging to Shared Channels (USCH/DSCH). † DL Power Timeslot Correction [TDD]: This function enables the Node B to apply an individual offset to the transmission power in each timeslot according to the downlink interference level at the UE.

Table 3.5 lists all the NBAP functions and the corresponding NBAP elementary procedures, utilised in the context of each function. Table 3.6 lists all the NBAP elementary procedures and the NBAP messages associated with each elementary procedure.

3.4.12 Establishment of Data Bearers in UTRAN Before finishing the UTRAN discussion, it is instructive to illustrate the way the data bearers (or transport channels) are established on a UTRAN interface. As an example, we will consider a case where a new RAB is established on the Iu-cs interface. Typically, on Iub and Iur interfaces all data bearers are based on AAL2 and are also called data streams. In order to distinguish the several data bearers existing on a particular interface (Iu, Iur, Iub), every data bearer features a unique radio-network identifier and a unique transportnetwork identifier on this interface. The radio-network identifier is the identity of the data

210

Broadband Wireless Mobile: 3G and Beyond

Table 3.5 NBAP functions and corresponding procedures NBAP Function

NBAP Elementary Procedures

Cell Configuration Management

X X X X

Common Transport Channel Management

System Information Management Resource Event Management

Configuration Alignment

Measurements on Common Resources

Radio Link Management

Radio Link Supervision Compressed Mode Control [FDD]

Measurements on Dedicated Resources

DL Power Drifting Correction [FDD]

Cell Setup Cell Reconfiguration Cell Deletion Common Transport Channel Setup

X Common Transport Channel Reconfiguration X Common Transport Channel Deletion X System Information Update X Block Resource X Unblock Resource X Resource Status Indication X Audit Required X Audit X Reset X Common Measurement Initiation X Common Measurement Reporting X Common Measurement Termination X Common Measurement Failure X Radio Link Setup X Radio Link Addition X Radio Link Deletion X Unsynchronised Radio Link Reconfiguration X Synchronised Radio Link Reconfiguration Preparation X Synchronised Radio Link Reconfiguration Commit X Synchronised Radio Link Reconfiguration Cancellation X Radio Link Pre emption X Radio Link Failure X Radio Link Restoration X Radio Link Setup X Radio Link Addition X Compressed Mode Command X Unsynchronised Radio Link Reconfiguration X Synchronised Radio Link Reconfiguration Preparation X Synchronised Radio Link Reconfiguration Commit X Synchronised Radio Link Reconfiguration Cancellation X Dedicated Measurement Initiation X Dedicated Measurement Reporting X Dedicated Measurement Termination X Dedicated Measurement Failure X Downlink Power Control

211

Network Architecture

Table 3.5 (continued) NBAP Function

NBAP Elementary Procedures

Reporting of General Error Situations Physical Shared Channel Management [TDD] DL Power Timeslot Correction [TDD]

X Error Indication X Physical Shared Channel Reconfiguration X Downlink Power Timeslot Control

bearer used by the radio signalling protocol (RANAP, RNSAP, NBAP), while the transportnetwork identifier is the identity of the data bearer used at the transport signalling protocol, or ALCAP. Table 3.7 shows the particular identities used to identify data bearer at the various UTRAN interfaces and at the various protocol layers. On the Iu interface, every data bearer is identified by a RAB-ID, while on Iur and Iub interfaces, every dedicated data stream is identified by a DCH-ID and every downlink shared data stream is identified by a DSCH-ID. Now let us assume that a new Iu-cs data bearer needs to be established. The establishment will initiate at the RANAP level and will include the following steps (see Figure 3.31). Step 1. The originating RANAP (o-RANAP) entity transmits a RAB-Assignment Request message to the terminating RANAP (t-RANAP) entity. Among others, the RAB-Assignment Request includes: † the RAB-ID (8 octets) of the requested data bearer, which is allocated by the o-RANAP entity; † the Transport Address, which should be used by the t-RANAP when requesting the new transport channel from the transport-network layer; † a Binding-ID (4 octets), which is another identifier used to associate the RAB-ID with the ALCAP-ID. The Binding-ID has significance only at the interface between the radionetwork layer and the transport-network layer.

Step 2. The t-RANAP processes the received RAB-Assignment request and, in response, requests from the t-ALCAP to establish the new transport channel. For this purpose, the Transport Address and the Binding-ID received by the o-RANAP are sent to the t-ALCAP. Step 3. Subsequently, t-ALCAP sends a request for a new transport channel establishment to the transport address specified by the Transport Address parameter. This request includes also the Binding-ID. Step 4. The o-ALCAP indicates to the o-RANAP that a new transport channel is requested and also indicates the Binding-ID associated with this request. Step 5. The o-RANAP will recognise this Binding-ID and will accept the establishment of this new transport channel. Step 6. Afterwards, o-ALCAP sends an accept message to t-ALCAP and also configures the user-plane at the originating side to connect the established transport channel to the correct UP protocols. Step 7. At this point, t-ALCAP configures the user plane at the terminating side and confirms to t-RANAP that the requested transport channel has been established. In turn, tRANAP responds to the o-RANAP with a RAB-Assignment Response message. From now on, o-RANAP (or t-RANAP) can release or modify this transport channel by

Radio Link Deletion

Common Measurement Initiation Radio Link Addition

Physical Shared Channel Reconfigure [TDD] Audit Block Resource Radio Link Setup System Information Update

Common Transport Channel Deletion

PHYSICAL SHARED CHANNEL RECONFIGURATION REQUEST AUDIT REQUEST BLOCK RESOURCE REQUEST RADIO LINK SETUP REQUEST SYSTEM INFORMATION UPDATE REQUEST COMMON MEASUREMENT INITIATION REQUEST RADIO LINK ADDITION REQUEST RADIO LINK DELETION REQUEST

CELL SETUP REQUEST CELL RECONFIGURATION REQUEST CELL DELETION REQUEST COMMON TRANSPORT CHANNEL SETUP REQUEST COMMON TRANSPORT CHANNEL RECONFIGURATION REQUEST COMMON TRANSPORT CHANNEL DELETION REQUEST

Cell Setup Cell Reconfiguration

Cell Deletion Common Transport Channel Setup Common Transport Channel Reconfiguration

Initiating message

NBAP elementary procedures and messages

NBAP elementary procedure

Table 3.6

CELL SETUP RESPONSE CELL RECONFIGURATION RESPONSE CELL DELETION RESPONSE COMMON TRANSPORT CHANNEL SETUP RESPONSE COMMON TRANSPORT CHANNEL RECONFIGURATION RESPONSE COMMON TRANSPORT CHANNEL DELETION RESPONSE PHYSICAL SHARED CHANNEL RECONFIGURATION RESPONSE AUDIT RESPONSE BLOCK RESOURCE RESPONSE RADIO LINK SETUP RESPONSE SYSTEM INFORMATION UPDATE RESPONSE COMMON MEASUREMENT INITIATION RESPONSE RADIO LINK ADDITION RESPONSE RADIO LINK DELETION RESPONSE

Response message on successful outcome

PHYSICAL SHARED CHANNEL RECONFIGURATION FAILURE AUDIT FAILURE BLOCK RESOURCE FAILURE RADIO LINK SETUP FAILURE SYSTEM INFORMATION UPDATE FAILURE COMMON MEASUREMENT INITIATION FAILURE RADIO LINK ADDITION FAILURE —

CELL SETUP FAILURE CELL RECONFIGURATION FAILURE — COMMON TRANSPORT CHANNEL SETUP FAILURE COMMON TRANSPORT CHANNEL RECONFIGURATION FAILURE —

Response message on unsuccessful outcome

212 Broadband Wireless Mobile: 3G and Beyond

Dedicated Measurement Reporting Dedicated Measurement Termination

Radio Link Restoration

Radio Link Failure

Audit Required Common Measurement Reporting Common Measurement Termination Common Measurement Failure Synchronised Radio Link Reconfiguration Commit Synchronised Radio Link Reconfiguration Cancellation

Synchronised Radio Link Reconfiguration Preparation Unsynchronised Radio Link Reconfiguration Dedicated Measurement Initiation Reset Resource Status Indication

RADIO LINK RECONFIGURATION PREPARE RADIO LINK RECONFIGURATION REQUEST DEDICATED MEASUREMENT INITIATION REQUEST RESET REQUEST RESOURCE STATUS INDICATION AUDIT REQUIRED INDICATION COMMON MEASUREMENT REPORT COMMON MEASUREMENT TERMINATION REQUEST COMMON MEASUREMENT FAILURE INDICATION RADIO LINK RECONFIGURATION COMMIT RADIO LINK RECONFIGURATION CANCELLATION RADIO LINK FAILURE INDICATION RADIO LINK RESTORE INDICATION DEDICATED MEASUREMENT REPORT DEDICATED MEASUREMENT TERMINATION REQUEST — — —

— — —

— —

—

— —

—

—

—

—

—

— —

RADIO LINK RECONFIGURATION FAILURE RADIO LINK RECONFIGURATION FAILURE DEDICATED MEASUREMENT INITIATION FAILURE — —

— —

RADIO LINK RECONFIGURATION READY RADIO LINK RECONFIGURATION RESPONSE DEDICATED MEASUREMENT INITIATION RESPONSE RESET RESPONSE —

Network Architecture 213

Error Indication Downlink Power Timeslot Control [TDD] Radio Link Pre-emption

DEDICATED MEASUREMENT FAILURE INDICATION DL POWER CONTROL REQUEST

Dedicated Measurement Failure Downlink Power Control [FDD] Compressed Mode Command [FDD] Unblock Resource

COMPRESSED MODE COMMAND UNBLOCK RESOURCE INDICATION ERROR INDICATION DL POWER TIMESLOT CONTROL REQUEST RADIO LINK PREEMPTION REQUIRED INDICATION

Initiating message

(continued)

NBAP elementary procedure

Table 3.6

— — —

—

—

— — —

—

—

— —

—

Response message on unsuccessful outcome

—

Response message on successful outcome

214 Broadband Wireless Mobile: 3G and Beyond

215

Network Architecture Table 3.7

Data bearer identifiers in UTRAN

Radio Network Signalling Protocol Identifier (RANAP/ RNSAP/NBAP ID) Transport Network Signalling Protocol Identifier (ALCAP ID)

Figure 3.31

Iu ps

Iu cs

Iur

Iub

RAB ID

RAB ID

DCH ID or DSCH ID

DCH ID or DSCH ID

IP Address 1 TEID

AAL2 VPI 1 VCI

AAL2 VPI 1 VCI

AAL2 VPI 1 VCI

Typical steps for establishing a new Iu cs data bearer.

sending suitable requests to o-ALCAP (or t-RANAP) and including the Binding-ID that corresponds to this channel.

3.5 Network Access Security In this section we discuss several aspects of the security of UMTS as specified in 3GPP specifications TS 33.102 [6], TS 21.133 [7] and TS 33.120 [8]. The main discussion is based on the general security architecture, as specified in [6]. This architecture is briefly explained and it is identified that one key component of this architecture is the network access security. The network access security is composed of a number of mechanisms aiming at satisfying the security requirements on the air interface. Such mechanisms include: † User identity authentication, i.e., verification that the user requesting service is indeed the one who claims he is. † User identity confidentiality, i.e., prevention of releasing any kind of identity that could help prospective eavesdroppers identify where a user is located and/or what type of services he utilises. † User data confidentiality, i.e., employment of data ciphering in order for this data to become meaningless to prospective eavesdroppers. † Integrity protection, i.e., verification that (1) data has not been changed by an unauthorised

216

Broadband Wireless Mobile: 3G and Beyond

user and (2) the entity that generated the data is the expected one and not an unauthorised one. Hence, by integrity protection we mean data integrity verification and origin authentication. † Network authentication, i.e., verification that the network, which announces that it provides 3G services, is an authorised one and can be trusted by the mobile. We should note that the latter two mechanisms, i.e. the integrity protection and the network authentication, are specific security enhancements of 3G systems and they are not provided in 2G systems. This section provides a thorough description of the above network access security mechanisms. In addition, it explores the security principles of UMTS and its security enhancements as compared to 2G systems. Finally, it identifies the main security objectives of UMTS as well as its key security principles.

3.5.1 Key Security Principles The security architecture of UMTS is based essentially on the following three key principles: 1. 3G security is built on the security of 2G systems. Security elements within GSM and other second generation systems that have proved to be needed and robust are adopted and implemented in 3G security architecture as well. 2. 3G security aims to improve the security of 2G systems. In this context, 3G security addresses many of the identified weaknesses of 2G systems and provides the means to mitigate or even eliminate these weaknesses. 3. 3G security aims to offer new security features and to secure new services offered by 3G platforms. The term ‘3G security’ in this section mainly refers to ‘UMTS security’. As implied by principle 1 above, several fundamental aspects of GSM security are retained in UMTS. In particular, the following security aspects of GSM are adopted in UMTS after some potential improvements: † Authentication of subscribers for service access: All subscribers are authenticated before granted the requested services. The authentication process aims to verify that the real identity of a subscriber is the same with the one claimed by the subscriber. In some cases, the authentication process is optional. UMTS security clarifies and tighten the conditions for optional authentication. † Radio interface encryption: Transmissions over the radio interface are encrypted to protect against eavesdrop by unauthorised individuals. In UMTS the encryption schemes intend to be more powerful to provide improved protection against sophisticated attacks. The encryption schemes have been designed having in mind the increased computing power that is available for cryptanalysis nowadays and in the near future. Moreover, UMTS security addresses the fact that the method of negotiating which encryption algorithm to be used is open to attack. † Subscriber identity confidentiality on the radio interface: Over the radio interface the permanent identity (i.e. IMSI) of a subscriber will be kept confidential and a temporary identity will be allocated by the network and will be used for identification purposes. UMTS aims at providing a more secure mechanism for allocating temporary identities.

Network Architecture

217

† USIM: A removable security module will be also used in UMTS to personalise a mobile terminal and to implement the security functionality in the user equipment. This module is still manageable by the network operators and programmed with network operator specific data. Also, this module is independent of the mobile terminal as regards its security functionality. † SIM application toolkit security features are retained for providing a secure application layer channel between the SIM/USIM and a home network server † The operation of security features is independent of the user, i.e. the user does not have to do anything for the security features to be in operation. † The home environment (HE) trust in the serving network (SN) for security functionality is minimised.

3.5.2 Weaknesses in Second-Generation Security Among other things, UMTS security architecture aims to eliminate the following weaknesses of GSM security (and other second-generation systems): † Active attacks using false BTSs are possible. † Cipher keys and authentication data are transmitted in clear between and within networks. † Encryption does not extend far enough towards the core network resulting in clear text transmission of user and signalling data across microwave links (in GSM, from the BTS to the BSC). † User authentication using a previously generated cipher key (where user authentication using RAND, SRES and A3/8 is not provided) and the provision of protection against channel hijack rely on the use of encryption, which provides implicit user authentication. However, encryption is not used in some networks, leaving opportunities for fraud. † Data integrity is not provided. Data integrity defeats certain false BTS attacks and, in the absence of encryption, provides protection against channel hijack. † IMEI is an unsecured identity and should be treated as such. † Fraud was not considered in the design phase of the second-generation systems but as afterthoughts to the main design work. † There is no HE knowledge or control of how an SN uses authentication parameters for HE subscribers roaming in that SN. † Second generation systems do not have the flexibility to upgrade and improve security functionality over time.

3.5.3 Security Objectives The general objectives of UMTS security can now be summarised as follows: 1. Ensure that information generated by or relating to a user is adequately protected against misuse or misappropriation. 2. Ensure that the resources and services provided by serving networks and home environments are adequately protected against misuse or misappropriation. 3. Ensure that the security features standardised are compatible and available worldwide (there shall be at least one ciphering algorithm that can be exported on a worldwide basis).

218

Broadband Wireless Mobile: 3G and Beyond

4. Ensure that the security features are adequately standardised to ensure worldwide interoperability and roaming between different serving networks. 5. Ensure that the level of protection afforded to users and providers of services is better than that provided in contemporary fixed and mobile networks. 6. Ensure that the implementation of UMTS security features and mechanisms can be extended and enhanced as required by new threats and services.

3.5.4 Security Architecture In each functional stratum of UMTS architecture [11] a number of security features are defined (see [6]). This is illustrated in Figure 3.32, where the arrows indicate interfaces over which security features are provided. Let us briefly discuss the security features provided over some interfaces. Between a mobile equipment (ME) and an access network (AN) the security features contain encryption of data and signalling, data integrity, etc. The arrow between the ME and the AN points both ways to emphasise that encryption and data integrity are provided bi-directionally. Between the USIM and the SN we have also bi-directional security features. In this case, the SN verifies that the USIM is valid and entitled to receive 3G services, while the USIM verifies that the SN is authorised to offer 3G services. The direction of the arrow between the USIM and the ME indicates that some information in USIM is kept confidential and should not be received by the ME. Hence, the information flow from the USIM to the ME must be governed by some security features. The security features between the USIM and the User indicate that only an authorised user should be capable to access USIM and that some USIM data should be protected from being accessed by the user. Finally, the arrow pointing from the HE towards the SN indicates that when the SN requests some security data from the HE, the latter should somehow verify that the former is a trusted network, which can receive the requested sensitive data. A simple method to categorise the security features is to take into account the security

Figure 3.32

UMTS security architecture.

Network Architecture

219

objectives and the threats associated with each feature. Based on that, we distinguish the following groups, which are also illustrated in Figure 3.32: † Network access security (I): The set of security features that provide users with secure access to 3G services and which, in particular, protect against attacks on the (radio) access link. † Network domain security (II): The set of security features that enable nodes in the provider domain to securely exchange signalling data and protect against attacks on the wireline network. The network domain security is typically up to the network operator or vendor. † User domain security (III): The set of security features that secure access to mobile stations. These features include:

User-to-USIM authentication: This ensures that access to USIM is restricted until the USIM has authenticated the user. Hence, it is ensured that access to the USIM can be restricted to an authorised user. To accomplish this feature, the user and the USIM must share a secret key (e.g. a PIN) that is stored securely in the USIM. The user gets access to the USIM only if he inputs the correct secret key. USIM-Terminal authentication: This ensures that access to a terminal is restricted to an authorised USIM. The USIM and the terminal must share a secret that is stored securely in the USIM and the terminal. In general, a USIM that fails to prove its knowledge of the secret, it will be denied access to the terminal. † Application domain security (IV): The set of security features that enable applications in the user and in the provider domain to securely exchange messages. Application domain security includes i)security mechanisms for accessing the user profile data and ii)IP security. In addition, it includes mechanisms to provide secure messaging between the network and the USIM, e.g. to protect messages transferred over the network to applications on the USIM. In the latter case, the following security features are provided:

Entity authentication of applications: two applications are able to corroborate each other’s identity. Data origin authentication of application data: the receiving application is able to verify the claimed data origin of the application data received Data integrity of application data: the receiving application is able to verify that application data has not been modified since it was sent by the sending application. Replay detection of application data: an application is able to detect that the application data that it receives is replayed. Sequence integrity of application data: an application is able to detect that the application data is received in sequence. Proof of receipt: the sending application can proof that the receiving application has received the application data sent. Confidentiality of application data: application data is not disclosed to unauthorised parties. † Visibility and configurability of security (V): The set of features that enables the user to inform himself whether a security feature is in operation or not and whether the use and provision of services should depend on the security feature.

220

Broadband Wireless Mobile: 3G and Beyond

From the above groups the one deserving special consideration is group (I), the network access security. This is because, in wireless networks, the access medium is inherently available to everyone and thereby it encounters the most attacks. For this reason, the network access security should be highly sophisticated and complicated to provide a high degree of protection. The following section discusses in more detail the security features provided in UMTS to realise the security requirements on the radio interface.

3.5.5 Network Access Security The network access security is effectively composed of the following security features: † † † †

user identity confidentiality entity authentication confidentiality data integrity.

3.5.5.1 User identity confidentiality This security feature aims at providing: † User identity confidentiality, which ensures that the permanent identity (IMSI) of a user to whom a service is delivered cannot be eavesdropped on the radio access link. † User location confidentiality, which ensures that the presence or the arrival of a user in a certain area cannot be determined by eavesdropping on the radio access link. † User untraceability, which ensures that an intruder cannot deduce whether different services are delivered to the same user by eavesdropping on the radio access link.

The main mechanism for providing user identity confidentiality is by identifying a user with a temporary identity or with an encrypted permanent identity. Below we explain how the temporary identity is used in a UMTS network. Identification by temporary identities Assume that a UE needs to attach to a UMTS network. For this purpose, it should send an Attach Request message, which should embed some kind of UE identity in order to make it possible for the network to perform identification and authentication procedures. The problem arising at this point is that the Attach Request message cannot be transmitted ciphered because some ciphering parameters (e.g. a ciphering key) need to be first negotiated between the UE and the network. For this reason, the messages transmitted by the UE before and during the ciphering negotiation must be transmitted unciphered. This fact makes the system vulnerable to malicious monitoring and specific mechanisms should be devised to provide an accepTable 3.level of protection against sensitive data, such as the user identity. To accomplish this goal, the network allocates a temporary identity to the UE, referred to as TMSI in the CS domain or P-TMSI in the PS domain. These temporary identities are used to identify the UE in the CS and in the PS domain respectively. For the purpose of the following discussion we will focus only on the PS domain and we will subsequently use only P-TMSI. We should keep in mind though that similar procedures apply for the CS

221

Network Architecture

Figure 3.33

Using P TMSI as a temporary user identity.

domain too. P-TMSI is unique within a specific routing area and is allocated either explicitly (with the P-TMSI Reallocation procedure, see [4]) or implicitly in the context of some mobility management procedures. P-TMSI is used as illustrated in Figure 3.33. Whenever a P-TMSI is allocated to an UE this value is stored in the USIM. Subsequently, when the UE sends an Attach Request it identifies itself with the P-TMSI value stored in USIM. If no valid P-TMSI is stored in USIM (e.g. at the very first Attach Request), the UE uses its permanent identity (IMSI) instead. In the example shown in Figure 3.33, the used P-TMSI value is denoted as P-TMSI0. This P-TMSI0 combined with the oldRAI (which is also included in the Attach Request message) points to a unique Mobility Management (MM) context in the network, which contains the UE’s permanent identity (IMSI). When the network (i.e. an SGSN in this example) responds to the Attach Request with an Attach Accept message, it allocates a new P-TMSI value to the UE, say, P-TMSI1. Note that the Attach Accept is typically transmitted in ciphered mode, hence, P-TMSI1 is protected against malicious monitoring. In this way, the new P-TMSI value allocated to the UE is considered to be known only to the UE and the network. The next time the UE will need to perform a mobility management procedure, it will use PTMSI1 to identify itself. For example, as shown in Figure 3.33, a subsequent Routing Area Update Request will include P-TMSI1. (In fact, the Routing Area Update Request does not explicitly contain the P-TMSI value, but this value can be derived by other means, which are not important in the current discussion.) Note that this request message is transmitted unciphered. In response, the SGSN will allocate a new P-TMSI value, P-TMSI2, to the UE within the Routing Area Update Accept message. In this way, the network allocates a new P-TMSI to the UE every time its previous one is transmitted unciphered. With the above mechanism, the UE uses P-TMSI to identify itself instead of its IMSI. It

222

Broadband Wireless Mobile: 3G and Beyond

should be noted that P-TMSI is only meaningful to the network, which contains an association between the allocated P-TMSIs and IMSIs. Typically, P-TMSI provides no information to a malicious user about the UE’s permanent identity. P-TMSI signature Suppose now that an unauthorised user sends a Routing Area Update Request using a P-TMSI value that is arbitrarily chosen. If it happens this P-TMSI value to be allocated to an authorised user, the network may erroneously believe that the Routing Area Update Request comes from the authorised user. Therefore, using only P-TMSI, one can provide identity confidentiality but may leave ‘open holes’ to the system for malicious attacks. To work around this security problem, the network will need to initiate an authentication procedure before processing the Routing Area Update request. However, this takes up valuable resources (e.g. additional messages sent to UE and possibly to HLR) and adds considerable delays. To avoid the frequent authentication procedures, the P-TMSI Signature may be used. PTMSI is an additional parameter with 3-octet length and is used as follows. Each time the SGSN allocates a new P-TMSI value to a UE, it also allocates a P-TMSI Signature that is associated with that P-TMSI. In the UE side, the P-TMSI Signature is stored in USIM together with P-TMSI, while, in the network side, both P-TMSI Signature and P-TMSI currently allocated to a UE are stored in the UE’s MM Context. Every time a UE sends a message that includes a P-TMSI value it should also include the associated P-TMSI Signature value. Hence the SGSN receives [P-TMSI, P-TMSI Signature] pairs. The SGSN checks if the received P-TMSI has been allocated and, if so, it checks the validity of this P-TMSI by means of the received P-TMSI Signature. If the received P-TMSI Signature is correct, the received P-TMSI is considered as valid; otherwise, it is considered invalid. In this way, the probability of an unauthorised user to choose a [P-TMSI, P-TMSI Signature] pair that is valid becomes negligible. Ciphering key sequence number When the P-TMSI Signature is used, a UE is authenticated by means of P-TMSI and PTMSI Signature and therefore there is no need to perform the normal authentication procedure. It is important to note that, even when this procedure is not performed, the ciphering of user data can still take place. The ciphering key stored at the mobile is identified to the network by means of a parameter called Ciphering Key Sequence Number (CKSN). In some UMTS specifications this parameter is also called as Key Set Identifier (KSI). In the example shown in Figure 3.33, the CKSN is included in the Routing Area Update Request sent by the mobile. From the received CKSN the network infers which was the last authentication vector used to authenticate the mobile and therefore identifies what is the Ciphering Key (CK) and the Integrity Key (IK) stored in the mobile. If these two keys match with the corresponding CK and IK stored in the network (i.e. in the MM Context of the mobile), then the network may start the ciphering and integrity check without invoking the authentication procedure. We must note however that all these hold true given that the MM Context of the mobile station is present in the network. If this MM Context does not exist, then the authentication procedure must be executed. On the other hand, if the CK and IK implicitly indicated by the

223

Network Architecture

mobile do not match with the corresponding values stored in the network (i.e. when the CKSN indicated by the mobile do not match with the CKSN stored in the network), then the network must authenticate the UE. 3.5.5.2 Entity authentication This security feature aims at providing: † User authentication, which ensures that the serving network confirms the identity of the user. † Authentication mechanism agreement, which ensures that the user and the serving network can securely negotiate the mechanism for authentication and key agreement that they shall use subsequently. † Network authentication, which ensures that the user confirms that he is connected to a serving network authorised by the user’s HE to provide 3G services; this includes the guarantee that this authorisation is recent.

Before we discuss in detail how the entity authentication is provided, we need to get familiar with the authentication vectors a special set of parameters used to authenticate an entity. Authentication vectors The information used to authenticate the user and the networkis composed of a number of socalled authentication vectors (see Figure 3.34). Each vector contains five authentication parameters: † † † † †

a random challenge RAND (128-bit long); an expected user response XRES (variable length); a cipher key, CK (128-bit long); an integrity key, IK (128-bit long); a network authentication token, AUTN.

Note that a GSM/GPRS authentication vector contains only the first three parameters. The purpose of the random challenge, RAND, is to provide the mobile station with a random number to be used to calculate the authentication response RES and both the ciphering key CK and integrity key IK. The same calculation is performed at the network side. The

Figure 3.34 UMTS authentication vectors.

224

Broadband Wireless Mobile: 3G and Beyond

Figure 3.35

The calculation of an authentication vector.

expected user response XRES is the response expected by the user after an authentication challenge. The actual user response RES is compared with the expected response XRES to determine whether the mobile, or more correctly, the USIM card, is authentic or not. The ciphering key CK is used for the ciphering of information, both at the network and at the mobile. Likewise, the integrity key is another key used at both the network and the mobile to provide data integrity. Finally, the authentication token AUTN is used to provide the mobile with a means to authenticate the network. Through the AUTN parameter, the mobile authenticates the HE, which supplied the authentication vectors. Figure 3.35 indicates how the various authentication parameters are calculated. Each time the network needs to generate a new authentication vector, fresh RAND and SQN values are generated. As shown in Figure 3.35, the random challenge RAND and the long-term secret key K are used as inputs to the authentication functions f2, f3, f4, which output XRES, CK and IK respectively. The long-term secret key K is generated when a new subscription is created. This key is never transmitted and is securely stored at the mobile station (in USIM) and at the authentication centre (AuC) in the network. The authentication functions f1-f5 shown in Figure 3.35 can be operator specific. They are executed by USIM when a mobile is challenged with a RAND value. In the network side they are executed in the AuC when new authentication vectors need to be calculated. As illustrated in Figure 3.35, the AUTN is the concatenation (denoted as ||) of three individual parameters: † the parameter derived by the exclusive OR between the sequence number, SQN, and the anonymity key, AK; † the authentication management field, AMF; † the message authentication code, MAC.

The SQN is a 6-octet sequence number generated before a new authentication vector is

Network Architecture

225

calculated. A method to generate SQN is specified in appendix C of [6]. The AK is an anonymity key used to conceal the sequence number, SQN, as the latter may expose the identity and location of the user. The concealment of the sequence number is to protect against passive attacks only. If no concealment is needed then f5 ; 0, AK 0. The AMF is a parameter that may be used to indicate the algorithm and the key used to generate a particular authentication vector. By means of AMF, a mechanism is provided to support the use of multiple authentication and key agreement algorithms. This mechanism is useful e.g. for disaster recovery purposes. The MAC parameter will be explained in a later section. Acquisition of authentication data Assume that an SGSN needs to authenticate a UE and assume that the SGSN maintains no authentication vectors for that UE. These authentication vectors may be acquired (1) from the UE’s HE (typically, an HLR), or (2) from another SGSN, which previously served the UE and has already acquired the necessary authentication information. The second method is typically more efficient, first, because it minimises long-distant accesses to the HE (in case of inter-PLMN roaming) and, second, because it prevents overloading of the HE and the corresponding bottlenecks. In addition, it may be considered faster when long-distant circuits are involved between the SGSN and the HE. When an SGSN needs to authenticate a UE and the authentication vectors cannot be retrieved from another SGSN, it sends a Send Authentication Info [IMSI] command to the appropriate HLR, which contains subscription information for that particular UE. Before sending the command the SGSN needs to know the IMSI of the UE. This IMSI value is used as a parameter to Send Authentication Info command and is used for deriving the SS7 address of the appropriate HLR. For roaming UEs, the HLR and the SGSN belong to different PLMNs. After receiving the command Send Authentication Info [IMSI], the HLR returns to the requesting SGSN a number of authentication vectors, typically from 1 to 5. The HLR either produces these vectors at the moment it receives the Send Authentication Info command (by sending another command to the AuC) or it simply retrieves these vectors from its storage, if these are already available (e.g. from a previous operation). Evidently, since an SGSN always retrieves authentication information from the user’s HE, the authentication procedure does not depend on the SN where the UE is located. Every time the SGSN needs to authenticate the UE, it uses an authentication vector, which has not been used before. If all the available authentication vectors have already been used, the SGSN will request a new set of authentication vectors from the HLR. Also, when the SGSN transfers the authentication vectors to another SGSN (e.g. during an inter-SGSN routing area update), it sends only the authentication vectors marked as not-used. We must note that in GSM an authentication vector is possible to be used more than once (the exact number is specified by the operator), however, in UMTS it is mandatory that each vector is used only once. Authentication procedure When the SGSN needs to authenticate a specific UE and already maintains a list of usable authentication vectors, it chooses an authentication vector from that list and sends to the UE an Authentication and Ciphering Request message that includes the RAND and AUTN values

226

Broadband Wireless Mobile: 3G and Beyond

Figure 3.36

Calculation performed by USIM after challenged by the network.

of the chosen authentication vector. Optionally, when a mobile is an old GSM mobile that does not support the UMTS authentication procedure, the network may not send the AUTN. In such cases, the mobile infers that the network performs a GSM authentication challenge (as opposed to the UMTS authentication challenge) and it will use the GSM-specific algorithms to calculate the authentication response and the ciphering key (in GSM, this key is denoted as Kc). Upon receiving the random challenge RAND and the AUTN, the USIM is expected to perform the calculation illustrated in Figure 3.36. Using the RAND and K, and by applying the authentication functions f2, f3, f4, the response RES, the ciphering key, CK, and the integrity key, IK, are derived respectively. Also, the expected MAC, XMAC, is calculated, as well as the sequence number. Subsequently, USIM verifies that the derived XMAC is equal to the received MAC and that SQN is in the allowed range. In such cases, the mobile considers the network as authentic and trusted for safe provision of 3G mobile services. Therefore, it responds to the authentication challenge with an authentication response message that includes the derived RES value. However, if XMAC is not equal to MAC, or if the SQN is not in the allowable range, the network authentication fails and the mobile responds with an authentication failure message. When the network receives the authentication response from the mobile it simply checks if the response, RES, matches with the expected response, XRES. In the case they do, the mobile is authenticated. However, when a mismatch between RES and XRES is identified, the USIM is not considered authentic and the network responds with an authentication reject message. The mechanism for providing authentication (of both user and network) and key agreement is depicted in Figure 3.37. This figure effectively illustrates the sequence of messages and events that take place in the context of this mechanism.

227

Network Architecture

Figure 3.37

Typical steps during authentication and key agreement.

In this example we assume that the mobile tries to attach to the PS domain of a serving network. Therefore, the MS transmits some kind of access request message, which is routed to the SGSN. (Equivalently, when the MS requests access to the CS domain, the access request is routed to the VLR.) Here the access request can be considered as a generic message corresponding to any kind of access request message, such as Attach Request, Routing Area Update Request, etc. The SGSN finds the IMSI of the MS, which is either included in the access request or is derived from the P-TMSI, as discussed previously. If the SGSN needs to authenticate the UE and has no authentication vectors for that UE, it needs to fetch these vectors from the HLR. As already mentioned, the HLR usually responds with 1 5 authentication vectors, which are stored at the SGSN. Subsequently, the SGSN uses an authentication vector, say with index i, to challenge the mobile. Note that the transmitted authentication request message, besides RAND(i) and AUTN(i), contains the Key Set Identifier (KSI), also referred to as CKSN. As illustrated in Figure 3.37, if the mobile and network authentication is successful (i.e. RES(i) is equal to XRES(i)), the ciphering mode is enabled and the SGSN updates the HLR with the new location of the mobile. If, however, the authentication fails the SGSN sends an Authentication Failure Report back to the HLR to report about this event. The UMTS network access security mechanism contains another enhancement as compared to the equivalent GSM security mechanism. In particular, the USIM contains a mechanism to limit the amount of data that is protected by a given set of ciphering and integrity keys, or to limit the time period wherein the same set of ciphering keys can be used. In other words, the USIM keeps track of the time period a given key set is being used and,

228

Broadband Wireless Mobile: 3G and Beyond

upon expiry, it may initiate itself an authentication procedure. This procedure follows the same principles discussed already, but could be triggered by the mobile station. 3.5.5.3 Setup of security mode Typically, at the beginning of a new signalling connection between the mobile and the network, the integrity protection and the confidentiality of signalling messages are enabled. When these security procedures are enabled, all subsequent signalling messages exchanged between the mobile and the network are subject to integrity protection and ciphering. Therefore, a signalling message at the receiving side (either the mobile or the network) is accepted only if its integrity is verified and if its origin is authenticated. Also, it will not be accepted if it has not been ciphered as expected. Since the entire network operation and the resources allocation are controlled by means of signalling messages, those messages are considered as rather sensitive and critical data. For this reason, it is generally desirable that all signalling procedures are executed with integrity protection enabled. However, some signalling procedures can be executed with no integrity protection. Such procedures include the identification by a permanent identity and the authentication and key agreement. The activation of integrity protection and ciphering is performed by the security mode setup procedure. Let us investigate the basic principles applied to the execution of this procedure. Before every signalling session between the mobile and the network, an RRC connection is established. When this connection is established, the mobile specifies to the RAN its security capability, which includes the UMTS Encryption Algorithms (UEAs) and the UMTS Integrity Algorithms (UIAs) supported by the mobile. Subsequently, the mobile sends its first NAS message to the CN, which can be, for example, a Location Update Request, a Routing Area Update Request, a CM Service Request, etc. This message includes the identity of the mobile (e.g. P-TMSI for the PS domain) and the Key Set Identifier (KSI), which, as we have seen already, is the KSI allocated by the CN during the last authentication procedure. At this point, the network may request to identify the mobile (i.e. request its IMSI) and perform again the authentication, or it may skip these procedures, if the reception of the correct KSI and P-TMSI (and possibly P-TMSI Signature) are considered as having enough authentication certificates. After that, the CN sends a Security Mode Command to the RAN in order to initiate integrity protection and ciphering. The Security Mode Command contains (1) a list of allowed UIAs and the IK and, optionally (if ciphering must be started), (2) a list of allowed UEAs and the CK. The RAN now knows the integrity and encryption algorithms supported by the mobile and the integrity and encryption algorithms allowed by the CN. In turn, it selects a preferable UIA and a preferable UEA. Obviously, these algorithms must be supported by the mobile and must be allowed by the CN. The RAN now generates a random value FRESH (see below), initiates the downlink integrity protection and sends a Security Mode command (an RRC message) to the mobile. This command specifies the UIA and FRESH values to be used for integrity protection and, if ciphering must be started, the UEA to be used for ciphering. It must be noted that, since the mobile could have two sets of ciphering and integrity keys (one for each CN domain), it is specified which of these sets must be used. Upon receiving the Security Mode command, the mobile verifies its integrity. The integrity check mechanism is described elsewhere in this section.

Network Architecture

229

If the integrity check is successful, the mobile initiates ciphering and integrity on the uplink and sends a Security Mode Complete to the RAN. Since integrity and ciphering have already been activated in RAN, the RAN will decipher the Security Mode Complete message and will verify its integrity by using the negotiated UIA and FRESH. The setup of security mode procedure ends with the RAN responding to the CN by a Security Mode Complete message (this is a RANAP message). It is important to understand that with the security mode setup procedure: † The downlink security starts with the Security Mode command sent on the downlink. This RRC command and all the following downlink RRC messages are integrity protected and possibly ciphered by the RAN. Consequently, the mobile performs integrity check and possibly deciphering to all downlink messages following the Security Mode command. † The uplink security start with the Security Mode Complete sent on the uplink. This RRC command and all the following uplink RRC messages are integrity protected and possibly ciphered by the mobile. Consequently, the RAN performs integrity check and possibly deciphering to all uplink messages following the Security Mode Complete command.

Having already discussed the security mode setup procedure, which specifies how integrity protection and ciphering is activated on the downlink and on the uplink, the two following sections will focus on explaining how ciphering and integrity checks are performed after they have been activated. 3.5.5.4 Confidentiality The confidentiality of user data and signalling is provided by means of ciphering. The following security features are provided for establishing confidentiality on the network access link: † cipher key agreement: the UE and the SN agree on a cipher key that they will use subsequently; † cipher algorithm agreement: the UE and the SN securely negotiate the cipher algorithm that they will use subsequently; † confidentiality of user data: user data cannot be overheard on the radio access interface; † confidentiality of signalling data: signalling data cannot be overheard on the radio access interface.

As already seen, the cipher key agreement is realised during the UE and network authentication procedure and results in the establishment of a common ciphering key (CK) at both the network and the UE (stored in USIM). Also, the cipher algorithm agreement is realised during the security mode negotiation, which has been discussed in a previous section. In the following, we will discuss how confidentiality of user data and signalling is provided by means of ciphering. Figure 3.38 illustrates the processing performed for ciphering (at the transmitting side) a plaintext block, i.e. a clear block of user data or signalling, and the processing performed for deciphering (at the receiving side) a ciphertext block, i.e. a ciphered block of user data or signalling. This processing takes place at the RLC layer and, therefore, the plaintext block is effectively an RLC block. As shown in Figure 3.38, the ciphering of the plaintext block is performed by XORing the plaintext block with a keystream block. The latter is derived at the

230

Broadband Wireless Mobile: 3G and Beyond

Figure 3.38

Ciphering/deciphering method

output of the encryption algorithm (UEA), which is generally denoted as function f8. The parameters applied to the input of f8 are CK, COUNT-C, BEARER, DIRECTION and LENGTH. At the receiving side, the same keystream block used to cipher a plaintext block is derived by applying exactly the same inputs to function f8. Finally, the original plaintext block is derived by XORing again the received ciphertext block with the keystream block. Below we briefly explain the input parameters to the ciphering function f8. More detailed information can be found in [6]. The ciphering key used to cipher a plaintext block can be either the CK corresponding to PS domain (CKps), or the CK corresponding to the CS domain (CKcs). Since the RLC layer at the mobile does not know whether CS or PS is the destination of a message, it cannot apply the ciphering key on a per-message basis. The general rule that governs the selection of the ciphering key is that the mobile will apply the ciphering key of the CN domain for which the most recent security mode negotiate took place. Hence, if the mobile is only attached to the CS domain, it will use CKcs. If, subsequently, it attaches to the PS domain and a security mode negotiation takes place with the PS domain, then the mobile will use CKps for all subsequent uplink messages (going either to CS or PS domain). COUNT-C is the ciphering sequence number and is 32 bits long. The update of COUNT-C depends on the transmission mode and the logical RLC channel used. For example, in RLC Acknowledged Mode, one part of the COUNT-C (12 bits) contains the RLC sequence number that is available in each RLC PDU and the other part (20 bits) contains the RLC Hyper-Frame Number [6]. Therefore, a different COUNT-C value is applied to every plaintext block. BEARER is the radio bearer identifier and is 5 bits long. There is one BEARER parameter per radio bearer associated with the same user and multiplexed on a single 10-ms physical layer frame. BEARER is input to the ciphering function f8 to avoid that for different keystream an identical set of input parameter values is used.

Network Architecture

231

DIRECTION is the direction identifier and is 1 bit long. The value of the DIRECTION is 0 for messages on the uplink and 1 for messages on the downlink. LENGTH is the length indicator and is 16 bits long. The length indicator determines the length of the required keystream block. LENGTH affects only the length of the keystream block but not the actual bits in it. Up to now, only two UEAs have been defined: UEA0, which corresponds to no encryption, and UEA1, which applies the so-called Kasumi algorithm. 3.5.5.5 Data integrity The following security features are provided with respect to integrity of data on the network access link: † integrity algorithm agreement: the UE and the SN can securely negotiate the integrity algorithm that they shall use subsequently; † integrity key agreement: the UE and the SN agree on an integrity key that they may use subsequently; † data integrity and origin authentication of signalling data: the receiving entity (UE or CN) is able to verify that signalling data has not been modified in an unauthorised way since it was sent by the sending entity (SN or UE) and that the data origin of the signalling data received is indeed the one claimed.

We have already seen that integrity key agreement is realised during the UE and network authentication procedure and results in the establishment of a common integrity key (IK) at both the network and the UE (stored in USIM). Also, the integrity algorithm agreement is realised during the security mode negotiation, which has been discussed in a previous section. In the following, we will discuss how integrity protection of signalling data is provided. It must be noted that, after an RRC connection establishment and after the security mode set-up, all dedicated control signalling messages (e.g. RRC, MM, CC, GMM, SM) are integrity protected. Figure 3.39 illustrates the method used to provide integrity protection and origin authentication. This method is realised at the RRC layer, as opposed to ciphering, which is realised at the RLC layer. The transmitting side transmits along with a message a kind of electronic signature, called Message Authentication Code (denoted as MAC-I). The receiving side processes the message and derives and the expected MAC value, XMAC, which is compared with the MAC sent by the originator. If the originator is the expected one, i.e. the one using the IK, FRESH and UIA previously negotiated, then, under normal conditions, the comparison of the expected and received MAC values will be successful and, therefore, the message will pass the integrity check and will be accepted. The input parameters to UIA, generally called integrity function f9, must be synchronised for the integrity protection to work correctly. The IK applied for the integrity protection can be either the one negotiated with the CS domain (IKcs) or the one negotiated with the PS domain (IKps). Since the RRC layer at the mobile does not know whether CS or PS is the destination of a message, it cannot apply the integrity key on a per-message basis. The general rule that governs the selection of the integrity key is the one mentioned for the ciphering key in the

232

Broadband Wireless Mobile: 3G and Beyond

Figure 3.39

Data integrity check and origin authentication method.

previous section. That is, the mobile will apply the integrity key of the CN domain for which the most recent security mode negotiate took place. † The MESSAGE applied to the input of f9 is the original RRC message, which has to be integrity protected. † COUNT-I is the integrity sequence number and is 32 bits long. One part of COUNT-I (4 bits) contains the RRC sequence number that is available in each RRC PDU and the other part (28 bits) contains the RRC Hyper-Frame Number [6]. Therefore, a different COUNT-I value is applied in every RRC message. † DIRECTION is the direction identifier and is 1-bit long. Its value is 0 for messages on the uplink direction and 1 for messages on the downlink direction. It is used to avoid computing message authentication codes for uplink and downlink messages with an identical set of input parameters to the integrity function f9. † FRESH is a random 32-bits parameter selected by the network during the security mode activation. There is one FRESH parameter value per user. As already seen, at connection setup the RAN generates a random FRESH value and sends it to the user in the RRC security mode command. This FRESH value is subsequently used by both the network and the user throughout the duration of a single connection. The input parameter FRESH protects the network against replay of signalling messages by the user. In other words, this mechanism assures the network that the user is not replaying any old MAC-Is.

Up to now, only one UIA has been defined: UEA1, which is also called Kasumi algorithm. 3.5.5.6 Ciphering and integrity protection in GPRS and UMTS Figure 3.40 illustrates the ciphering and integrity protection mechanisms into the protocol architecture of GPRS and UMTS. As shown, in GPRS only ciphering is provided, which is

233

Network Architecture

Figure 3.40

Ciphering and integrity protection in GPRS and UMTS

performed at the LLC layer. On the other hand, UMTS provides ciphering at the RLC layer, both in the user plane and the control plane. In addition, UMTS provides at the RRC layer (i.e. at the control plane) integrity protection of signalling messages. This integrity protection enhances the robustness of UMTS and its protection against malicious attacks. Note that integrity protection is not provided at the user plane, since user data is typically not interpreted by the network and therefore is not so critical to the system operation. 3.5.5.7 Abbreviations 2G 3G 3GPP AAL5 AK AKA AMF AN APN ATM AUTN AUTN AV BG BSSAP1 BSSGP

Second Generation Third Generation Third Generation Partnership Project ATM Adaptation Layer type 5 Anonymity Key Authentication and key agreement Authentication management field Access Network Access Point Name Asynchronous Transfer Mode Authentication Token Authentication Token Authentication Vector Border Gateway Base Station System Application Part 1 Base Station System GPRS Protocol

234

Broadband Wireless Mobile: 3G and Beyond

BVCI BSSGP Virtual Connection Identifier CGF Charging Gateway Functionality CK Cipher Key CKSN Cipher key sequence number CMM Circuit Mobility Management CS Circuit Switched DHCP Dynamic Host Configuration Protocol DNS Domain Name System GEA GPRS Encryption Algorithm GERAN GSM EDGE Radio Access Network GGSN Gateway GPRS Support Node GMM/SM GPRS Mobility Management and Session Management GSN GPRS Support Node GTP GPRS Tunnelling Protocol HE Home Environment HLR Home Location Register IETF Internet Engineering Task Force IK Integrity Key IMSI International Mobile Subscriber Identity IP Internet Protocol IPv4 Internet Protocol version 4 IPv6 Internet Protocol version 6 KSI Key Set Identifier LAI Location Area Identity LLC Logical Link Control MAC Medium Access Control MAC Message Authentication Code MAC A The message authentication code included in AUTN computed using f1 ME Mobile Equipment MS Mobile Station MSC Mobile Switching Centre NSAPI Network layer Service Access Point Identifier PDN Packet Data Network PDP Packet Data Protocol, e.g. IP PDU Protocol Data Unit PLMN Public Land Mobile Network PPP Point to Point Protocol PS Packet Switched P TMSI Packet TMSI PTP Point To Point RA Routing Area RAB Radio Access Bearer RAC Routing Area Code RAI Routing Area Identity RANAP Radio Access Network Application Protocol RAND Random challenge RAU Routing Area Update RLC Radio Link Control RNC Radio Network Controller RNS Radio Network Subsystem RNTI Radio Network Temporary Identity RRC Radio Resource Connection SC Short Message Service Centre SGSN Serving GPRS Support Node SIM (GSM) Subscriber Identity Module

Network Architecture

SM SN SNDCP SQN SRNC SRNS TCAP TCP TFT TLLI TMSI UDP UE UEA UIA URA USIM UTRAN VLR XRES

235

Short Message Serving Network SubNetwork Dependent Convergence Protocol Sequence number Serving RNC Serving RNS Transaction Capabilities Application Part Transmission Control Protocol Traffic Flow Template Temporary Logical Link Identity Temporary Mobile Subscriber Identity User Datagram Protocol User Equipment UMTS Encryption Algorithm UMTS Integrity Algorithm UTRAN Registration Area User Service Identity Module UMTS Terrestrial Radio Access Network Visitor Location Register Expected Response

3.5.5.8 Definition of terms Access Control: The process that makes sure that unauthorised use of a resource is prevented. Authentication: The process that makes sure that the claimed identity of an entity is correct. Integrity: The process that makes sure that information has not been changed by unauthorised parties. Cloning: The process of changing the identity of one entity to that of an entity of the same type, so that there are two entities of the same type with the same identity. Confidentiality: The property that information is not made available or disclosed to unauthorised individuals, entities or processes. Data integrity: The property that data has not been altered in an unauthorised manner. Data origin authentication: The corroboration that the source of data received is as claimed. Entity authentication: The provision of assurance of the claimed identity of an entity. Key freshness: A key is fresh if it can be guaranteed to be new, as opposed to an old key being reused through actions of either an adversary or authorised party. USIM User Services Identity Module. In a security context, this module is responsible for performing UMTS subscriber and network authentication and key agreement. It should also be capable of performing GSM authentication and key agreement to enable the subscriber to roam easily into a GSM Radio Access Network. SIM GSM Subscriber Identity Module. In a security context, this module is responsible for performing GSM subscriber authentication and key agreement. This module is not capable of handling UMTS authentication nor storing UMTS style keys. UMTS Entity authentication and key agreement: Entity authentication according to this specification.

236

Broadband Wireless Mobile: 3G and Beyond

GSM Entity authentication and key agreement: Entity authentication according to TS ETSI GSM 03.20 User access module: either a USIM or a SIM Mobile station, user: the combination of user equipment and a user access module. UMTS subscriber: a mobile station that consists of user equipment with a USIM inserted. GSM subscriber: a mobile station that consists of user equipment with a SIM inserted. UMTS security context: a state that is established between a user and a serving network domain as a result of the execution of UMTS AKA. At both ends ‘UMTS security context data’ is stored, that consists at least of the UMTS cipher/integrity keys CK and IK and the key set identifier KSI. GSM security context: a state that is established between a user and a serving network domain usually as a result of the execution of GSM AKA. At both ends ‘GSM security context data’ is stored, that consists at least of the GSM cipher key Kc and the cipher key sequence number CKSN. Quintet, UMTS authentication vector: temporary authentication data that enables an VLR/SGSN to engage in UMTS AKA with a particular user. A quintet consists of five elements: a) a network challenge RAND, b) an expected user response XRES, c) a cipher key CK, d) an integrity key IK and e) a network authentication token AUTN. Triplet, GSM authentication vector: temporary authentication data that enables a VLR/ SGSN to engage in GSM AKA with a particular user. A triplet consists of three elements: a) a network challenge RAND, b) an expected user response SRES and c) a cipher key Kc. Authentication vector: a vector comprising of five parameters in UMTS (referred also as authentication quintet) or three parameters in GSM/GPRS (referred also as authentication triplet). Temporary authentication data: either UMTS or GSM security context data or UMTS or GSM authentication vectors. References and Further Reading [1] Otsu, T., Okajima, I., Umeda, N., Yamao, Y., ‘Network architecture for mobile communications systems beyond IMT 2000’ IEEE Pers. Commun., Vol. 8, No. 5, pp. 31 37, Oct. 2001 [2] 3rd Generation Partnership Project, TS 23.127: ‘Virtual Home Environment (VHE); Stage 2’. [3] 3rd Generation Partnership Project, TS 23.002: ‘Network Architecture.’ [4] 3rd Generation Partnership Project, TS 23.060: ‘General Packet Radio Service (GPRS); Service Description; Stage 2.’ [5] GSM 03.20: ‘Digital Cellular Telecommunications System (Phase 21); Security Related Network Functions.’ [6] 3rd Generation Partnership Project, TS 33.102: ‘Security Architecture.’ [7] 3rd Generation Partnership Project, TS 21.133: ‘3G Security; Security Threats and Requirements’ [8] 3rd Generation Partnership Project, TS 33.120: ‘3G Security; Security Principles and Objectives’ [9] 3rd Generation Partnership Project web site, www.3gpp.org [10] 3rd Generation Partnership Project, specifications archive: http://www.3gpp.org/ftp/ specs [11] 3rd Generation Partnership Project, TS 23.101: ‘General UMTS Architecture.’ [12] P.K. Bhatnagar,Engineering Networks for Synchronisation CCS7, 1997, 0 7803 1158 2. [13] 3rd Generation Partnership Project, TS 23.110, ‘UMTS Access Stratum; Services and Functions’. [14] 3rd Generation Partnership Project, TS 25.401, ‘UTRAN Overall Description’. [15] 3rd Generation Partnership Project, TS 25.412, ‘UTRAN Iu interface Signalling Transport’. [16] 3rd Generation Partnership Project, TS 25.422, ‘UTRAN Iur interface Signalling Transport’. [17] 3rd Generation Partnership Project, TS 25.432, ‘UTRAN Iub interface Signalling Transport’.

Network Architecture

237

[18] 3rd Generation Partnership Project, TS 25.414, ‘UTRAN Iu Interface Data Transport & Transport Signalling’. [19] 3rd Generation Partnership Project, TS 25.424, ‘UTRAN Iur Interface Data Transport & Transport Signalling for CCH Data Streams’. [20] 3rd Generation Partnership Project, TS 25.434, ‘UTRAN Iub Interface Data Transport & Transport Signalling for CCH Data Streams’. [21] 3rd Generation Partnership Project, TS 25.426, ‘UTRAN Iur and Iub Interface Data Transport & Transport Signalling for DCH Data Streams’. [22] 3rd Generation Partnership Project, TS 25.413, ‘UTRAN Iu Interface RANAP Signalling’. [23] 3rd Generation Partnership Project, TS 25.423, ‘UTRAN Iur Interface RNSAP Signalling’. [24] 3rd Generation Partnership Project, TS 25.433, ‘UTRAN Iub Interface NBAP Signalling’. [25] 3rd Generation Partnership Project, TS 25.427, ‘UTRAN Iur and Iub Interface User Plane Protocols for DCH Data Streams’. [26] 3rd Generation Partnership Project, TS 25.425, ‘UTRAN Iur Interface User Plane Protocols for CCH Data Streams’. [27] 3rd Generation Partnership Project, TS 25.435, ‘UTRAN Iub Interface User plane Protocols for CCH Data Streams’. [28] 3rd Generation Partnership Project, TS 25.331, ‘Radio Resource Control (RRC) Protocol Specification’. [29] 3rd Generation Partnership Project, TR 25.933, ‘IP Transport in UTRAN’. [30] 3rd Generation Partnership Project, TS 25.304, ‘UE Procedures in Idle Mode and Procedures for Cell Reselec tion in Connected Mode’. [31] 3rd Generation Partnership Project, TS 23.122, ‘NAS Functions Related to Mobile Station (MS) in Idle Mode.’ [32] ITU T Recommendation Q.711 (07/1996): ‘Functional Description of the Signalling Connection Control Part’. [33] ITU T Recommendation Q.712 (07/1996): ‘Definition and Function of Signalling Connection Control Part Messages’. [34] ITU T Recommendation Q.713 (07/1996): ‘Signalling Connection Control Part Formats and Codes’. [35] ITU T Recommendation Q.714 (07/1996): ‘Signalling Connection Control Part Procedures’. [36] ITU T Recommendation Q.715 (07/1996): ‘Signalling Connection Control Part User Guide’. [37] ITU T Recommendation Q.2100 (07/1994): ‘B ISDN Signalling ATM Adaptation Layer (SAAL) Overview Description’. [38] ITU T Recommendation Q.2110 (07/1994): ‘B ISDN ATM Adaptation Layer Service Specific Connection Oriented Protocol (SSCOP)’. [39] ITU T Recommendation Q.2140 (02/1995): ‘B ISDN ATM Adaptation Layer Service Specific Co ordination Function for Signalling at the Network Node Interface (SSCF AT NNI)’. [40] ITU T Recommendation Q.2210 (07/1996): ‘Message Transfer Part Level 3 Functions and Messages Using the Services of ITU T Recommendation Q.2140’. [41] ITU T Recommendation I.363.5 (08/1996): ‘B ISDN ATM Adaptation Layer Specification: Type 5 AAL’. [42] ITU T Recommendation I.363.2 (09/1997): ‘B ISDN ATM Adaptation Layer Specification: Type 2 AAL’. [43] ITU T Recommendation Q.2630.1 (12/99): ‘AAL type 2 Signalling Protocol (Capability Set 1)’. [44] ITU T Recommendation Q.2150.1 (12/99): ‘B ISDN ATM Adaptation Layer Signalling Transport Converter for the MTP3b’. [45] G. Sidebottom, et al., SS7 MTP3 User Adaptation Layer (M3UA), IETF Internet Draft. [46] R. R. Stewart, et al., Stream Control Transmission Protocol, IETF Internet Draft. [47] J. Loughney, et al., SS7 SCCP User Adaptation Layer (SUA), IETF Internet Draft. [48] 3rd Generation Partnership Project, TS 25.304, ‘UICC terminal Interface; Physical, Electrical and Logical Test Specification’. [49] RFC 2543: ‘SIP: Session Initiation Protocol’. [50] 3rd Generation Partnership Project, TS 23.228, ‘IP Multimedia Subsystem (Stage 2)’. [51] 3rd Generation Partnership Project, TS 23.218, ‘IP Multimedia (IM) Session Handling; IP Multimedia (IM) Call Model’. [52] 3rd Generation Partnership Project, TS 24.228, ‘Signalling Flows for the IP Multimedia Call Control Based on SIP and SDP’. [53] 3rd Generation Partnership Project, TS 23.071, ‘Location Services (LCS); Stage 2’. [54] 3rd Generation Partnership Project, TS 43.051, ‘GSM/EDGE Radio Access Network (GERAN); Overall Description Stage 2’. [55] 3rd Generation Partnership Project, TS 29.002, ‘Mobile Application Part (MAP) Specification’.

238

Broadband Wireless Mobile: 3G and Beyond

[56] 3rd Generation Partnership Project, TS 23.205, ‘Bearer independent Circuit switched Core Network; Stage 2’. [57] A. K. Salkintzis, ‘Wireless IP with GPRS: Fundamental Operational Aspects,’ Invited Paper, 4th IEEE Int, Symp. Wireless Personal Multimedia Communications (WPMC’01) pp. 9 12, 2001. [58] A. K. Salkintzis, ‘Wide Area Wireless IP Connectivity with the General Packet Radio Service,’ Wireless IP, Artech House, in press.

4 Emerging Wireless Applications and Protocols 4.1 Introduction Two drivers have taken the market by surprise in recent years: wireless is the first and the second is the Internet and Internet technology. At the end of 1998 there were around 260 million wireless mobile users and growth continues at over ten million new subscribers per month. The latest forecast figures by Delson Group indicate that the total number of wireless cellular subscribers world-wide will reach two billion by 2005. Convergence of wireless and Internet will be one of the hottest topics in the coming years. Wireless Mobile Internet (or WMI) and its devices are coming to the market in ever increasing numbers all over the world. The investment in this business will be over 100 billion dollars within the next several years. The various sectors of the industry have also been active on the standards front and in the creation of alliances and forums that will carry the momentum forward. The WAP Forum and the WAP (Wireless Application Protocol) standards are particularly significant since this industry standard insulates the Internet-based application from the wireless mobile network infrastructure. Thus, content (service and application) can now be decoupled from delivery (bear network). This allows the applications to run over wireless mobile networks, both current and future, thereby creating a massive new global market for wireless value-added services. Wireless mobile users will not just want to access all the information that is available on the Internet via a phone. It therefore makes better sense if the operator selects an appropriate service portfolio that is tailored to mobile subscribers. It is also impractical to deliver the content to a phone in the same manner as that supplied to a PC. The graphics and hyperlinks need to be stripped out, but this only represents a Band-Aid approach to the problem. What’s needed is a wireless equivalent to the WWW, which is where the WAP comes into the picture. WAP is an industry standard that allows software developers to create content suitable for display on ‘smart’ mobile phones. They write their service according to the WAP specification and it can then be delivered over any mobile network. Developers do not need to know

240

Broadband Wireless Mobile: 3G and Beyond

anything about the delivery process and writing WAP-compliant applications is very similar to creating WWW content. Therefore, WAP hides the complexity of wireless networks, both current and future, at the application layer. This means that it is very easy to develop lightweight Internet applications that will run on the new wireless information devices. In addition to the WAP standards, there are also several other solutions available in the market, for example, iMode and wJAVA (wireless JAVA). If anyone has any doubt about the potential of wireless Internet, they need only look at the iMode service provided by the Japanese service provider NTT DoCoMo. In only about 15 months, DoCoMo built its iMode service into a popular wireless experience serving nearly 10 million users. Currently, a large amount of information exists in the Internet that can meet very diverse needs. This is clearly a good match for mobile wireless users while travelling from place to place. It has been projected that most handsets will be equipped with WAP and Wireless Java in a few years for wireless Internet applications. When Internet information can be delivered wirelessly according to handset geolocation information, it becomes even more valuable and convenient.

4.2 Wireless Application Protocol (WAP) The significance of WAP goes far beyond merely superposing the Internet on wireless networks. The WAP business will be one of the next storms in the coming years.

4.2.1 WAP Markets Why is the WAP so important? It is driven by the huge wireless mobile market as well as the Internet infrastructure. From Figure 4.1, the number of mobile phones will reach around 1.3 billion by 2004, and most of the phones will be WAP equipped. The importance of WAP is reflected in the following statements: † Mobile will make the Internet ubiquitous. Most of the applications are things that we already do the industry’s task is to convert a share of our use onto wireless Internet.

Figure 4.1

Worldwide Internet subscribers, 1999 2004 by IDC.

Emerging Wireless Applications and Protocols

Figure 4.2

241

Wireless Internet and WAP Market (source: OVUM).

† Wireless Internet is a strategic imperative. Operators must provide these services to avoid revenue and profit erosion. † WAP is a key building block for wireless Internet.WAP has received huge industry backing; it opens the market to new players.

Therefore, the wireless Internet will be big and WAP will be a main contender (see Figure 4.2 by OVUM). By March 2001, over 40 million WAP devices had been shipped worldwide (Figure 4.3), and thousands of applications are already available in the market including (but not limited to): † † † † † † † † †

M-commerce shopping, ticket purchases, reservations, comparison shopping Finance statements, funds transfer, shares trading M-billing notification, presentation and payment of bills Enterprise access inventory, shipment/sales updates, email access M-care customer service, payment status, other backroom operations Entertainment games, gambling, interactive multi-player events Messaging communication and collaboration Travel scheduling, advisories, reservations Location-smart services traffic reports, parking information, store discounts, event recommendations In the global market:

† Over 150 carriers deployed or in final testing of WAP. † Tens of thousands of developers are now creating applications and contents.

242

Broadband Wireless Mobile: 3G and Beyond

Figure 4.3

WAP devices worldwide (Source: WAP Forum).

† Over 12,000 WAP sites are available from 100 countries. † Over 5 million WAP-readable pages available. † Over 40 million WAP-enabled handsets are in circulation worldwide.

As shown in Figure 4.2, the revenue from WAP-like active micro-browser users is growing very fast (Figure 4.4). By 2005, it will reach over $300 billions. In a word, WAP is an application tool towards wireless Internet that meets:

Figure 4.4

Revenue from Active Micro Browser Users (source: OVUM).

Emerging Wireless Applications and Protocols

243

† anytime, any place and anywhere access † more mobile terminal users than PC users † mobile terminals are personal devices

However, the key success factors for WAP depend on the followings: † † † †

service marketing price time to access service overall user experience and service design

4.2.2 WAP Architectures and Protocols The focus of WAP is to bring mobile data to the mass market. The way of providing this is to map internet-type applications plus telephony-specific features onto wireless requirements. Internet pages known from the fixed network will be processed by compressing information and representing this in so-called ‘cards’ on the handset display. A very important advantage of WAP is that it aims to provide standardised support for wireless data applications. So each handset developed according to the WAP standard is able to support WAP-based applications in the whole network. With this technology in place, a new information service can simply be developed as a web site and then viewed on the screen of the phone. 4.2.2.1 What is WAP? WAP is the set of technical standards that enables users of wireless devices to access, and receive content from the Internet with: † † † † †

open, non-proprietary global device independent bearer independent a full-fledged Internet citizen

WAP is a standard developed by consensus from engineers, scientists, specialists and experts from over 500 companies all around the world. It adopts Internet protocol to the requirements of the wireless networks (see Figure 4.5). WAP also opens the world of mobile data services, for example, it brings mobile data to the mass market, Internet-type applications plus telephony-specific features as well as providing standardised support for wireless data applications. WAP is also applicable to many different wireless beares and infrastructures (see Figures 4.6 and 4.7). The major WAP units and functions include: WAP gateway † serves as proxy † provides protocol mapping between standard and WAP protocol

244

Broadband Wireless Mobile: 3G and Beyond

Figure 4.5

WAP overview (source: Siemens).

† provides encoding/decoding for efficient transfer of data † provides access to mobile data bearers

WAP servers † supports standard Internet protocols † provides application contents and scripts in WAP-specific or standard Internet formats

Figure 4.6

WAP applicable to many wireless beares (source: Siemens).

Emerging Wireless Applications and Protocols

Figure 4.7

245

WAP infrastructure (source: Siemens).

WAP clients † † † † †

supports WAP protocol stack provides encoding/decoding enables browsing of WAP contents enables user intersection supports intersection with SIM

WTA Server † † † †

located within secure network operator domain communicates with client via gateway able to ‘Push’ WAP contents to WAP client Figure 4.8 shows this structure.

4.2.2.2 WAP Roadmap The goal of WAP forum is to make the mobile device a first class citizen of the web. WAP fills a unique role of: † intersection of mobile/web/Internet † wireless industry focal point, for example, requirements, ideas, design and development as well as interoperability † collaborator with key web/Internet groups, for example, technical development, advisor on wireless capabilities and requirements † industry forum for multi-vendor interoperability

The WAP roadmap is shown in Figure 4.9. The focus includes the following.

246

Broadband Wireless Mobile: 3G and Beyond

Figure 4.8

WAP units and functions (source: Siemens).

Bandwidth Adaptation and evolution of the protocol suite: † 2.5G and 3G bearers: † convergence with IP environment (for example, TCP, TLS, …).

Creation of new services, enabled by bandwidth: † streaming media † larger data types (for example, animations)

Figure 4.9 WAP Roadmap.

Emerging Wireless Applications and Protocols

247

Convergence Collaboration with IETF, W3C and others: † to define a common Internet framework † suitable for mobile and fixed devices

Never ending process: † WAP 1.X leveraged HTTP, XML, URLs and the web † WAP 2.X will expand to leverage XHTML, TCP, etc.

Joint work in the past includes: † XHTML-basic † CC/PP/UAProf [ ]

Future work: † protocol optimisations (for example, TLS, HTTP, TCP) † application models (for example, voice, multi-mode UI)

Multimedia Mutimedia services enabled by: † † † † † † †

more bandwidth (GPRS, 3G, etc) Moore’s law and handset evolution imaging colour, standard formats, etc. audio download clips, streaming, etc. video download clips, streaming, etc. streaming protocols longer term, QoS and isochronous protocols

Advanced user-interface Handset evolution drives requirements: † pixel resolution, density and colour † new input models (voice, pointers, security sense, etc.)

Application model enhancements: † † † † †

styling model document model advanced script integration voice interface muti-mode user interface

Application services Additional application services enabled by terminal and network capabilities with Webenabling unique wireless features:

248

† † † †

Broadband Wireless Mobile: 3G and Beyond

terminal location push commerce capabilities messaging

Management services Services targeted at network operator: † provisioning † terminal monitoring † terminal management

4.2.2.3 WAP Traffic Model The WAP traffic model takes into account a couple of input parameters like: † user behaviour (WAP usage time, BH portion, view time, etc) † different applications and their usage † application characteristics (transactions, bandwidth volume)

And provide us the information about: † traffic per use † overall bandwidth demand (upstream/downstream) † number of transactions to WAP devices (NetServer, firewall, WAP server, WAP gateway)

Figure 4.10 shows a simple WAP traffic model. The WAP traffic model is only for general-purpose analysis of the WAP performance, and should not be taken as a standard model in different applications and services. 4.2.2.4 WAP and XML XML is a framework for markup languages, but not a markup language itself. WAP is the only wireless Internet protocol that is built on XML. Therefore, XML is a platform technology, not a presentation language. XML allows content to be published one time, but can be used many times. However, WML (the core markup language of WAP) is designed for presentation. The background of XML is: † XML is a platform technology:

Basis for many markup languages: SMIL, MathML, WML, XHTML, VoiceXML, etc. Basis for much application middleware and protocols. † XML is not a presentation language like HTML. † W3C has based a family of languages on XML:

XHTML next generation HTML (migrated to the XML platform); SMIL synchronised multimedia integration language; MathML mathematical markup.

Figure 4.10

A simple WAP traffic model (source: Siemens).

Emerging Wireless Applications and Protocols 249

250

Broadband Wireless Mobile: 3G and Beyond

Hence, WAP is a very early adopter of XML as WML, and many WAP applications use XML as a platform to implement application-specific middleware. This technology helps lead to device independence. Figures 4.11 and 4.12 show the typical XML usage in web applications. The future markup directions include: † Markup language convergence. † Convergence platform is XHTML:

Figure 4.11

Typical XML Usage

I (source: WAP Forum).

Figure 4.12

Typical XML Usage

II (source: WAP Forum).

Emerging Wireless Applications and Protocols

251

XHTML is next-generation version of HTML based on XML; XHTML has mudularity, enabling diverse uses. † Multiple profiles built on XHTML (for example, TV, Phone, PDA, PC).

In a word, WAP is ready for advanced features, for example, graphics, colour, animation, large file downloads, multimedia. 4.2.2.5 Next generation of WAP The next generation of WAP will include: † † † † † † † † † †

XHTML TCP colour graphics animation large file downloading location-smart services streaming media data synchronisation with desktop PIM advanced capabilities and features open, non-proprietary device and bearer independent

There are also lots of other issues related to the next generation WAP, for example, eventbased billing, transaction, SNMP support, intelligent routing as well as performance optimisation.

4.2.3 WAP Securities The wireless Internet demands very high security requirement, especially for banking, financing and mobile commerce. These and hundreds of others are basing their future applications on WAP. The initial WAP is secure, because: † WAP provides transport-level security (WTLS) (see Figure 4.13); † works for most commerce applications; † implicit trust of gateway operator comparable to trusting voice switch operator.

Figure 4.13 WAP transport level security (source: WAP Forum).

252

Broadband Wireless Mobile: 3G and Beyond

Figure 4.14

WAP proxy navigation (source: WAP FORUM).

Meanwhile, the proxy navigation enhances the WAP security (see Figure 4.14): † handset temporarily uses an alternative (secure) proxy; † secure proxy is located within for example, bank or corporate firewall; † packets between the handset and the secure proxy remain encrypted at all times.

WAP is by design, highly secure for mobile and location-smart (for example, store discounts and order, event booking) commerce, and the transactions are as secure as PC sites. By mid of 2001, the WAP security will meet the most extreme demands of total encryption and secure proxies in handsets and gateways which is suitable for 3Gwireless deployments. In the future 3Gwireless or 4Gmobile phones, security will be even more important. The user fingerprinting function (for example, sensor and processor) will be equipped. The phone can be multi-functional and re-configurable for different persons. In the application layer, WAP handles end-to-end direct security signalling to ensure the secure transmission.

4.2.4 WAP Interoperability Interoperability is an essential topic for communications, containing clear specifications, good developer guidance and education, a reference pool of products and well-established conformance and certification processes. The certification process insures the interoperability in that: † † † † †

starts with content and authoring guidelines uses a reference pool of products done by independent, third parties objective and confidential optimised from time to time

Hence, the WAP certification process includes full interoperability and compliance program covering the entire value chain (see Figure 4.15) which is comprised of two phases: † Application layer testing:

designed to provide both interoperability and compliance testing; product is tested against other products; comprehensive test suite is used.

Emerging Wireless Applications and Protocols

Figure 4.15

253

WAP certification (source: WAP Forum).

† Protocol layer testing:

leverages the application layer testing; protocol is tested against the specification; designed to ensure the device complies with the standard at the protocol layer. The WAP testing is the foundation (on compliance), but not the building, the WAP vendor testing is still required for WAP interoperability. The future WAP interoperability will focus on: † Universal accessibility:

ubiquitous access to Internet content and services from all wired and wireless clients. † Widen the scope of the Internet architecture:

enable access to content best suited for the requesting client. † Ensure interoperability and backward compatibly.

The universal accessibility defines the concepts of: † Publish once, render differently:

content published in a neutral manner; presentation and content are separated. † Rendering is specific to client needs:

narrow the rendering gap focus on a core mark-up; provide modules to serve specific domains, for example, phones, lap-tops, TV browsers, etc. † Deliver content using any available bearer:

narrow bearer, high-speed bearer; IP convergence or gateway/proxy as needed. † Provides access to the content regardless client or network capabilities.

254

Broadband Wireless Mobile: 3G and Beyond

Figure 4.16

An Overview of Universal Accessibility (source: WAP Forum).

Figure 4.16 illustrates an overview of these concepts. From the technical points of view, it introduces: † Native content:

content written in an abstract XML format; describes content as objects; allows content to be published once. † User-agent profiles:

provides a means to identify the client capabilities; declares mark-up needs, input/output needs; enables servers to render for the client. † Style sheets:

define graphical representation of the content; content server renders based on client and application needs.

4.2.5 WAP and 3Gwireless Third generation wireless (3Gwireless) will be the next storm in communications. 3Gwireless offers: † good IP data bearer † better ‘base’ bandwidth (for example, 32K or better) † scalable bandwidth on demand

However, the spectrum is still finite, and spectrum efficiency is very important to the carrier business model as well as the user cost model. 3Gwireless (including ongoing GPRS) and WAP are excellent partners:

Emerging Wireless Applications and Protocols

255

† Bandwidth helps deliver rich content:

without an application layer, it’s a pipe; no standard, no market expansion. † Bandwidth drives innovation

but standards drive interoperability.

Also, WAP and 3Gwireless network is a perfect match, because: † Bandwidth is always good:

more bandwidth richer service platform; new innovation and specification opportunities for example, multimedia; existing WAP services run better/faster/cheaper. † ‘All-IP Net everywhere’:

simplifies application creation; economics of scale in infrastructure. † Bandwidth is a clear driver of innovation:

allows WAP to add new features and services. As an ongoing solution towards 3Gwireless, GPRS is becoming popular. What enables users to exploit the WAP experience on GPRS? The answer is: new handsets and mobile commerce. † New handsets:

will be new data oriented models; peaceful co-existence: keypad, screen pointer, security check and voice recognition; prediction: a raft of better handsets from both traditional and a new wave of handset manufacturers. † Mobile commerce:

WAP is a key enabler; source of significant future revenues; new stuff: bar code scanning, electronic wallet, e-passport, etc. On the other hand, 3Gwireless needs WAP in that: † † † †

3Gwireless provides superior wireless network layer Networks require an application model WAP is a mobile application model WML browsers are prerequisite and optimised for wireless unique ergonomics small screens provision of location information telephony integration transaction nature, not browsing pushed info

256

Broadband Wireless Mobile: 3G and Beyond

Figure 4.17 3Gwireless WAP Phones (source: NTT DoCoMo).

In summary, 3Gwireless and WAP will work together to deliver the best services to the future mobile users: † better networks and better devices offer unique opportunities for evolution; † feature evolution in WAP drives traffic/usage; † interoperability ensured by WAP.

Figure 4.17 illustrates some 3Gwireless WAP phones.

4.2.6 WAP Services and Applications WAP allows all kind of applications which can be customised per user’s demands: Information † † † † † † †

web browsing news weather, sports, events, tickets product info, commercial offers booking and reservation public information / broadcast location dependent information (yellow pages, tourist info)

Financial services † e-payment

Emerging Wireless Applications and Protocols

257

† online banking † online shopping † auctions

Entertainment † † † † † †

virtual stores music on demand games on demand video clips lottery / gambling e-postcard

Office information † † † † †

intranet access news virtual working groups groupware schedule synchronisation mobile office

Communication † e-mail, messaging † personal organiser † remote expertise

Machine-to-machine services † telematic services:

traffic services navigation assistance vehicle tracking (GPS) fleet management † security monitoring services:

household device control † telemetry:

traffic monitoring remote sensing Table 4.1 also lists some applications from the other point of view. As an example, on-line mobile banking will have huge markets in the coming years (see Figure 4.18). Online Banking is based on the current electronic commerce that is offered on the present day Internet. The user should be able to perform all applications where he was urged in the

258

Broadband Wireless Mobile: 3G and Beyond

Table 4.1 Potential WAP applications (source: OVUM). Information Content

News flash, sports, stock quotes, event listings

Entertainment

Games, horoscopes, jokes, music downloads, video clippings Banking, auctions, ticketing, purchasing, loyalty, bill payment Coupons, discounts, sponsored services, directories PIM, e mail/messaging, instant messaging, unified messaging, chat rooms Intranet and database access, groupware, file transfer Tracking, dispatch, field sales automation Navigation and mapping Remote control, monitoring, configuration

Mobile e commerce Advertising Personal communications Mobile office Business processes Telematics Telemetry

past to enter a bank office. Electronic commerce may be defined as ‘any form of business transaction in which parties interact electronically rather than by physical exchanges or direct physical contact’. Electronic commerce is technology for change. Analysts expect the Internet e-commerce to have nearly ten-times more transactions by the year 2004 in comparison to 2000. The attractiveness of this application is based on the fact that a bank can save up to 95% of its personnel compared to a customer visiting office and the user is not confined by fixed opening hours and the geographical distribution of the branches.

Figure 4.18

On line mobile banking (source: Siemens).

Emerging Wireless Applications and Protocols

259

The necessary equipment is identical with the one from the information services, although a security layer for achieving authentication, encryption and privacy has to be added to guarantee safe transactions. In the future mobile phone (3Gwireless and 4Gmobile), the video screen will get bigger and bigger while the keypad will disappear. The following functions will be enhanced: † wireless keypad or keyboard with Bluetooth (or other personal access) technology as an option; † information recognition, for example, voice recognition, pattern recognition; † security detection, for example, fingerprinting sensing; † bandwidth-on-demand with peer-to-peer direct signalling.

WAP will address the detailed application layer protocol to ensure these functions are working properly. In addition, the future mobile applications will address the following issues: Location and orientation dependent applications and services Location awareness will play a major role for future wireless services. Currently, there are many projects working in this field, mostly about virtual tourist guides and similar scenarios. However, to exploit the full range of possibilities, major issues have yet to be solved. Micropayment system for content integrated in wireless systems Mobile content services and electronic commerce are expected to form a large part of valueadded services for future mobile phone systems. There is a great need and potential for mobile terminals to become ‘electronic wallets’ and media access devices for a large number of citizens. A major issue to be solved is payment and charging. Current charging systems for mobile phones are mainly based on network mechanisms, i.e. connection time, transfer of amount of messages/blocks/bits, or transaction-based charging. Also, service bundles and monthly subscription fees are coming in the same manner as in the Internet. What is missing clearly is service-based charging. For example on-demand-content, such as MP3 audio, on-demand-movies, mobile games and entertainment are very potential services. The use patterns for mobile services can be very scattered, so people may use these services only short times during their travel and empty slots in their daily work, etc. Micropayment is defined as a small payment, less than one dollar, which is too small to be charged immediately, and too big to be ignored. Micropayments are end-to-end visible, like money, so the receiver and payer can see, understand and trace their comings and goings. Potentially, large numbers of these payments need to be handled by the network and content producers, banks and customers. Micropayments are discrete in time, but a stream of micropayments can be used to charge continuous streaming type of service. Micropayments can be positive (charging) or negative (earn by using a service), routed (someone else is paying the cost, for example, sponsored service). Micropayments can be composite, so they can accumulate costs from lower levels of ingredients, for example a MP3 piece of music can be charged with micropayments assigning revenue directly to the musicians.

260

Broadband Wireless Mobile: 3G and Beyond

‘Open Source’ paradigm adoption especially for 3rd party APIs The open source paradigm means that: † a cumulating body of source code is available to the developers free of cost; † a co-ordination function is put into place which tests and accepts the increments to the body; † voluntary developers participate to the development of applications based on their own business, learning or hobby motivations.

During recent years the open source software development paradigm has brought up surprisingly good results, such as the emergence of the Linux operating system, Apache server, and GNU compilers. Besides superior quality and innovativeness, this paradigm has brought a large number of young people into the development of new applications all over the world, thus heavily contributing to the business process. Besides software development, IC and processor core designs have been influenced by this approach recently. Mobile communications have traditionally been divided into manufacturer/operator proprietary part, and into public interfaces defined in standards. Typically the proprietary part is where the differentations of the products and services are performed, giving a competitive edge to the manufacturers or service provider in question. However, there are new classes of services that are very difficult to specify and implement by a single manufacturer or operator. For example, mobile games form a quite promising area where evolutionary development by the users and enthusiasts may create a high volume of services. With the rising number of services and the convergence of IP and cellular networks, automatic service discovery will be a very important feature in future wireless network scenarios. With service discovery, devices may automatically discover network services including their properties, and services may advertise their existence in a dynamic way. WAP will help configure this new feature in the application layer. In summary, future generation wireless terminals will face significant challenges with respect to their ability to adapt to all kinds of options. In addition to the WAP improvement, it is important to continue research in the area of wireless transmission technologies, focusing on broadband components, signal processing and software definable radio.

4.2.7 WAP System Solutions The WAP concept provides the operator with a powerful environment to offer the subscribers value-added services which will boost the usage of data. The concept is realised by introducing a new type of node in the network, called the WAP Gateway. This node is able to distribute all WAP traffic to the correct application server. Figure 4.19 shows an overview of the WAP system framework. For an example, to provide WAP within the GSM/GPRS infrastructure existing today, the following additional elements are necessary: † WAP Gateway this is an element located at the wireless network operator’s domain which provides the bridge between wireless networks and the Internet:

serves as proxy which fetches the requested data from Internet sites;

Emerging Wireless Applications and Protocols

Figure 4.19

261

Generic overview of the WAP system framework (source: Siemens).

provides protocol mapping between standard and WAP protocol; provides encoding/decoding for efficient transfer of data bearers. † Mobile Data Application Server is located at the wireless operator’s site. It runs applications provided by the mobile operator and serves as an additional source of revenues. This server is:

located within secure network operator domain; communicates with client via gateway; able to ‘push’ WAP contents to WAP client; evolution towards Wireless Telephony Application Server (WTA) which is not yet finally standardised but provides the possibility to initiate a mobile-originated call setup directly from the WAP applications. † WAP Servers run by external service providers and make the whole world of information and services of the Internet accessible for subscribers using mobile phones equipped with a WAP browser. † Mobile Terminal WAP clients, have to include a WAP browser in order to support WAP based mobile services. These terminals:

support WAP protocol stack provide encoding/decoding enable browsing of WAP contents provide execution environment enable user interaction support interaction with SIM WAP system solution has been a hot topic since 2000, and lots of issues are related to the 3Gwireless network architecture, for example, All-IP signalling and protocol, networking topology, dynamic bandwidth allocation, access control and quality of services.

262

Broadband Wireless Mobile: 3G and Beyond

4.3 i-Mode 4.3.1 What is i-Mode? First introduced in Japan in February 1999 by NTT DoCoMo, i-mode is one of the world’s most successful services offering wireless web browsing and e-mail from mobile phones in Japan. The number of subscribers topped 20 million on March 4, 2001. Currently, 828 companies are providing information services through i-mode. This is a twelve-fold increase from the 67 firms that were signed on with the service at the time of its launch. In addition, there are about 1480 official i-mode Web sites including 59 i-applicompatible sites along with another 40,053 independent sites, according to OH! NEW isearch, an i-mode search engine. Since it is based on packet-data transmission, users do not pay for the time they are connected to a website or service. They are charged according to the volume of data transmitted. Unlike WAP, which uses WML (wireless markup language) as its markup language, imode services are built using i-mode-compatible HTML. Figure 4.20 shows the basic i-Mode concept in Japan. When you focus on the number of Internet users for private use, it is far from ‘critical mass’ for major IP in Japan. However, if you can count on all cellular phone users of DoCoMo, IP can easily reach economies of scale for its own service.

4.3.2 i-Mode-Compatible HTML i-Mode-compatible HTML is based on a subset of HTML 2.0, HTML 3.2 and HTML 4.0 specifications that was extended by NTT DoCoMo with tags for special use on cell phones, such as the ‘tel:’ tag, which is used to hyperlink a telephone number and let users initiate a call by clicking on a link.

Figure 4.20

Basic i Mode Concept: The Mobile ISP and Portal (source: MPT of Japan).

Emerging Wireless Applications and Protocols

263

I-mode-compatible HTML websites are easy to navigate, since all basic operations can be performed using a combination of four buttons: cursor forward, cursor backward, select, and back (return to previous page). Most phone browsers also provide a ‘Page Forward’ button. Functions that require two-dimensional navigation, such as image maps, and functions that require more intensive processing, such as frames and tables, are not included in the standard i-Mode HTML specifications. Here is an overview of some major features of HTML that are not included in i-Modecompatible HTML: † † † † † † †

JPEG images tables image maps multiple character fonts and styles background colours and images frames style sheets

Note that some handsets do support features beyond the standard cHTML specifications, such as tables and multiple fonts. But since these are not part of the published spec, you should use them with care and at your own risk! Since i-Mode-compatible HTML is based on standard HTML, developers can make use of and adapt millions of HTML-based content resources, various software tools, and public materials (textbooks, magazines, and web information).

4.3.3 i-Mode Network Structure The network structure of i-Mode is seen in Figure 4.21. The protocol stack is shown in Figure 4.22 where PDC (Japan’s 2G mobile systems) is for the wireless air link in the lower layer. As the 3Gwireless is taking off, the bandwidth will be increased to provide more attractive services.

Figure 4.21

Network structure of i Mode (source: DoCoMo).

264

Broadband Wireless Mobile: 3G and Beyond

Figure 4.22

Network Protocol of i Mode (source: DoCoMo).

4.3.4 Features of i-Mode The success of i-Mode comes from the following features: † Packet-based telecommunication charge (PDC-P) (see Figure 4.23):

PDC-P network enables cheaper communication charge for limited text-based data exchange; Mobile phone is always ‘stand by’ to receive data. † HTML-based

the world standard (see Figure 4.24):

Subset of HTML 3.0 with some additional tags;

Figure 4.23

Packet Based i Mode (source: DoCoMo).

Emerging Wireless Applications and Protocols

Figure 4.24

265

HTML Based i Mode (source: DoCoMo).

No need to learn special language to deliver content. In addition, there are lots of other features enhanced, for example, security, quality of services.

4.3.5 i-Mode Applications With i-Mode, the subscriber can: † Reserve airline and concert tickets, check their bank balances or transfer money, etc. † Send and receive e-mails. (The system is compatible not only with other i-Mode users, but also with PC and PDA users. Subscribers will be supplied with an e-mail address that consists of their cellular phone number followed by @docomo.ne.jp). † Access the Internet directly from their i-Mode-compatible cellular phone. † Register for Message Service, a specialised service which automatically provides subscribers with information on weather, depending on choice. † As i-Mode is based on a packet-data (9600 bps) transmission system, subscribers will be charged according to the volume of data transmitted, not the time spent on line.

Figure 4.25 illustrates a service image of i-Mode. It provides a customised homepage driven by user profile to make easier user accessibility to Internet web sites. i-Mode’s applications cover everywhere, i.e. mobile banking service, mobile trade service, credit card information, insurance transaction, travel information, airline and concert ticket reservation, news, entertainment, sales as well as corporate Intranet service. After 3Gwireless is launched world wide in 2002, the i-Mode picture will be much more beautiful and promising.

266

Broadband Wireless Mobile: 3G and Beyond

Figure 4.25

Service Image of i Mode (source: DoCoMo).

4.3.6 i-Mode Developing Strategy Success of i-Mode will lead to smooth introduction of 3Gwireless which enables 64 384 kbps high speed access from mobile devices. Even before 3Gwireless, DoCoMo made a weighty effort for enhancing the i-Mode platform by introducing Intranet package solutions in 1999, then Java technology, in 2000 and 2001. Figure 4.26 shows this developing strategy. The possible evolutions include: † i-Mode evolution 1: terminal development

new features for richer content

coloured LCD display (256 colours, 108 by 96 dots up) 4 grayscale large display poliphonic ringing tone MIDI format less weight more standby time new power technology † i-Mode evolution 2: Java-capable phone

Java can enable application/content providers to distribute software (applets) to cellular phone. You will be able to download applets such as interactive games, client agent software to manage servers, or security enhancing module and so on. Commercialisation would be in late 2000 and 2001 Figure 4.27 shows an example of agent type applet in i-Mode phone (base phone is Panasonic P501i).

Emerging Wireless Applications and Protocols

267

Figure 4.26 I Mode Developing Strategy (source: DoCoMo).

In the near future, the 3Gwireless-equipped i-Mode phone will bring new life to the wireless communications (see Figure 4.28): † wideband CDMA will allow faster download speeds and will provide richer content for iMode; † Video streaming, audio streaming, and much more …

4.4 Other Wireless Mobile Internet Application Technologies In addition to the i-Mode Java project by NTT DoCoMo in cooperation with Sun Microsystems, etc., Wireless Java (wJava) has become one of the hottest topics in the wireless

Figure 4.27

Java capable i Mode phone.

268

Broadband Wireless Mobile: 3G and Beyond

Figure 4.28

Next generation i Mode

3Gwireless equipped (source: DoCoMo).

mobile Internet technologies. The Mobile TCP/IP is still very cool in the industries especially when the IPv6 progresses very well. Lots of 3Gwireless technical forums (i.e. 3GPP, MWIF) are also working on the high layer application protocols to be applied for the high-speed wireless Internet services. With the development of wireless mobile Internet technology more and more solutions will come out; this will continue to be a very hot topic in the field.

4.5 Conclusions The year 2001 was the start of the wireless revolution for the new century. The Internet and Internet technology, as used everywhere all over the world, took the world by storm. The Internet Protocol (IP) became the lingua franca of data communications and it was becoming hard to imagine a business life without e-mail and instant access to information. This info phenomenon has now reached students and other young people, which means that the Wireless Mobile Internet is also becoming a consumer marketplace. In 1990, people dreamed of a mobile phone. By 2003, we shall be taking it for granted to own a wireless communicator with full wireless mobile Internet services anytime, anywhere and anyone. The biggest developments are 3Gwireless and WAP, and their impact will be ubiquitous. WAP will allow software developers to write applications and create services for a new, global market. By the year of 2005, there will be over 1.5 billion users of mobile phones. Wireless will not only be a technology, it will be a part of life. The service providers and network operators will be able to host and manage these new services in a completely different business model, therefore generating additional revenue and allowing offers to be clearly differentiated. When the 3Gwireless is in place it will

Emerging Wireless Applications and Protocols

269

become a brand new ball game. The smart wireless phones (WAP or wJava) will always be on-line; bandwidth-on-demand; highly secured; voice recognised; and users will pay on the basis of real usage, not time and distance. The wireless mobile Internet is therefore rapidly becoming a reality and the future looks very bright. References and Further Reading [1] [2] [3] [4] [5]

M. Mackenzie, et al., ‘A Presentation to the WAP Forum’, Feb. 5, 2001. R. Conway, ‘WAP in the GSM World’, Feb. 2001, WAP Forum Meeting, London. S. Goldman, ‘WAP: Fact and Future’, Mar. 2001, WAP Forum. WAP and the New Developments, Nov. 2000, Siemens AG. The Book of Visions 2000 Visions of the Wireless World, European Commission IST 2000. [6] DoCoMo’s i Mode Towards Mobile Multimedia in 3G, May 2000, NTT DoCoMo. [7] End to End IP Based Mobile Networks, Siemens AG. [8] Value added Services in Cellular Networks, CMG Telecommunications.

WSI Project, Nov.

5 Initiatives in 4Gmobile Design 5.1 Introduction – Who Needs 4G? What is 4G? 5.1.1 Social Background and Future Trends There has been an evolutionary change in mobile communication systems every decade. The increase in the number of subscribers and transmission data rates leads to a shift to higher frequency bands where wider bandwidth is available. There are two directions for future trends in mobile communications. One is Direction A, which aims to increase transmission data rates where the same mobility as IMT-2000 from indoor to high speed vehicles is maintained. The other is Direction B, which aims to expand mobility from indoor to outdoor where the transmission data rates in the wireless access systems are maintained. Direction B will be suitable for spot area services in order to satisfy the demand for higher data rates, while Direction A will accommodate continuous area services. We focus on Direction A here. The number of subscribers using PDC and PHS in Japan was 62.2 million in October 2000 and has increased by 10 million for 5 successive years. One in 2.6 mobile phone users connected to the Internet in October 2000. This pattern of use has increased significantly, and Internet use is expected to comprise 90 percent of all mobile communication traffic in 2005. The US and Europe are expected to follow a similar trend. Figure 5.1 shows a traffic forecast for Region 3 [97]. From 1999 through 2010, subscribers to voice-oriented services are expected to grow by 1.5 times, and the ratio between voice and multimedia traffic will be nearly 1:2 for total up- and downlinks. Assuming that multimedia traffic grows by 40 percent a year after 2010, it will be 23 times that of 1999, and the ratio between voice and multimedia traffic will be about 1:10. Therefore, to accommodate the considerable multimedia traffic after 2010, we must conduct R&D on key technologies to achieve not only high speed mobility but also high transmission data rates.

5.1.2 Trends in ITU-R At the 18th TG-8/1 in November 1999, the standardisation activities on IMT-2000 were finished and the new working party (WP8F) was established to co-ordinate on systems beyond IMT-2000 as well as to enhance IMT-2000 itself. At the first WP8F meeting in

272

Broadband Wireless Mobile: 3G and Beyond

Figure 5.1

Traffic Forecast for Region 3 in 2010 and after.

March 2000, 6 working groups were set up (see Figure 5.2) and their terms of reference are as follows [22]. WG-Vision: † Provision of the roadmap to the future in relation to the time perspectives for IMT-2000 and systems beyond IMT-2000. † Co-ordinate and complement the near term aspects of the Radio Technology Working Group and co-ordinate with the other groups in WP 8F. † Conceptualisation of the longer term future (5 to 10 years) and migrate it through a middle defining stage (3 to 7 years) to ultimately deliver a near term work product of specifications as defined in related working groups. † Maintenance and update of other IMT-2000 recommendations (such as concepts, principles, framework requirements and the like).

WG-Circulation: † Address to issues that may facilitate the ability of IMT-2000 to achieve global deployment including access, circulation, and common emission requirements.

WG-Developing IMT: † Consideration of issues relevant to the needs of the developing countries. † Assurance of the work on IMT-2000 adequately reflects these needs. Conduct of studies in response to Question ITU-R 77/8 and strengthening the liaison with ITU-D as necessary. † Maintenance and update of relevant IMT-2000 recommendations as appropriate may occur within this working group.

Initiatives in 4Gmobile Design

273

Figure 5.2 Working Group relationship diagram.

WG-Radio technology: † Maintenance and update of IMT-2000 RSPC terrestrial component in conjunction with external organisations. † Maintenance and update of IMT-2000 RSPC satellite component in conjunction with WP 8D. † Maintenance and update of other IMT-2000 recommendations as appropriate may occur within this working group; address aspects of adaptive antennas for IMT-2000 including technical characteristics, advantages, performance implications and applications. † Consideration of other aspects of technology related to IMT-2000; reception of the work products of the ‘mid-term’ perspective of the WG-Vision and in the ‘near-future’ updates existing specification recommendations or develops new recommendations as appropriate to support implementations of these concepts. † Co-ordination with external organisations in this task will be required.

WG-Spectrum: † Spectrum matters related to IMT-2000 and systems beyond IMT-2000; considering IMT2000 spectrum implementation issues and any necessary sharing, compatibility and interference criteria between IMT-2000 and other radio services

274

Broadband Wireless Mobile: 3G and Beyond

Figure 5.3 Target of Service beyond IMT 2000.

† Maintenance and update of existing IMT-2000 spectrum related recommendations and reports, as appropriate † Identifying areas where joint work and/or liaison is needed on spectrum matters with other relevant groups, as appropriate.

WG-Satellite Co-ordination: † Act as internal WP 8F co-ordinating function and focal point for satellite aspects † Function as the WP 8F point of interface for draft liaison statements to WP 8D on satellite issues † Maintenance and update of ITU-R recommendations related to IMT-2000 and systems beyond IMT-2000 and will work closely with WP 8D † Determine which WP 8F documents are relevant to WP 8D.

According to the WG-Vision, a target of service beyond IMT-2000 is illustrated with that of each mobile communication and wireless access service in Figure 5.3, and four scenarios for the systems beyond IMT-2000 are proposed, as shown in Figure 5.4 [23]. Scenario 1: All-round-type (Figure 5.4(a)) Covers the whole range of the deployment area and the transmission rate.

Initiatives in 4Gmobile Design

275

Figure 5.4 Scenarios of systems beyond IMT 2000: (a) all round type, (b) complement type, (c) area complement type, and (d) rate complement type.

Scenario 2: Complement-type (Figure 5.4(b)) Located in the position not covered by the other mobile communication and wireless access systems, both in terms of the deployment area and the transmission rate. Scenario 3: Area-complement-type (Figure 5.4(c)) They cover the whole range of the transmission rate and are located in the position not

276

Broadband Wireless Mobile: 3G and Beyond

Figure 5.4 (continued)

covered by the other mobile communication and wireless access systems in terms of the deployment area. Scenario 4: Rate-complement-type (Figure 5.4(d)) They cover the whole range of the deployment area and are located in the position not covered by the other mobile communication and wireless access systems in terms of the transmission rate.

277

Initiatives in 4Gmobile Design

Figure 5.5

Mobile communications systems.

5.1.3 Wireless Access Systems Related to 4G Mobile The following system requirements should be met in 4G mobile: † † † † †

high data-rate transmission high mobility wide coverage area and seamless roaming among different systems higher capacity and lower bit cost wireless QoS resource control

Because it is especially hard to realise the system having both high data rates and high mobility, four scenarios are discussed in WG-Vision (see section 5.1.2). There is an idea to include new communication systems such as ITS (Intelligent Transport Systems) and HAPS (High Altitude stratospheric Platform Station) systems at a research level (see Figure 5.5). There is also another approach organised in a layered structure similar to hierarchical cell structures in cellular mobile systems (see Figure 5.6), where vertical handover between systems as well as horizontal handover within a system is necessary [83].

5.1.4 Key Technologies It is very important to develop key technologies realising high data rates transmission under high mobility. Some of them are illustrated in Figure 5.7 [97], and recent research activities are introduced in detail after 5.2.

5.2 Microwave Propagation In recent years there has been a tremendous upsurge in demand for terrestrial mobile wireless

278

Broadband Wireless Mobile: 3G and Beyond

Figure 5.6

Layered structure of seamless future network [83].

communications, yet the availability of spectrum for mobile communications has become increasingly scarce as the information-oriented society has continued to evolve. In pursuit of new spectrum for mobile communications, many are casting their eyes to the hopeful prospects of the microwave band [105]. In order to adopt the microwave spectrum for use by mobile communications, it is essential to gain a clear understanding of the propagation

Figure 5.7

Key technologies in wireless access networks.

Initiatives in 4Gmobile Design

279

characteristics of the frequency band and bandwidth of the spectrum to be used. While the sub-microwave band (,2 GHz) has already been extensively developed [4,10], the propagation characteristics of microwave (.3 GHz) transmission at 10 Mb/s or more continue to be studied and still have numerous issues to be addressed. Microwave transmission exhibits greater path losses than conventional UHF-band communications, and it is assumed this can be attributed to frequency selective fading associated with broadband transmission and increased shadowing caused by shrinking Fresnal zones. In order to identify potential service areas, it is essential to clarify attenuation caused by multiple rays as a function of distance, attenuation caused by obstructions, time variation of multiple rays, and so on. In addition, there are other issues that must be addressed if we are to implement high-speed, high-quality microwave transmission including: delay time difference characteristic of multiple ray propagation paths and the problem of multiple rays separating and combining. Note too that demand and environments of interest are no longer confined, as they once predominately were, to cities. As the number of SOHO (small office, home office) workers and home-based telecommuters continues to grow, the service demand area is being rapidly pushed out into suburban residential areas, so careful thought must be given to the impact that this kind of environment has on microwave propagation. The only way to resolve these issues is to gather the relevant data through well-designed experimental studies, and to develop accurate models based on the data.

5.2.1 Microwave Mobile Propagation Characteristics in Urban Environments 5.2.1.1 Propagation loss characteristics LOS propagation losses and breakpoint characteristics Microcellular systems featuring transmitting base station antennas installed at a height of 4 m show excellent promise for transmitting in the microwave band. The attenuation coefficient of line-of-sight (LOS) path loss characteristics for these systems can be divided into a 2nd power domain and a 4th power domain, and results for propagation characteristics along roads is in agreement with existing reports for lower frequency transmissions. This point of transformation is referred to as the breakpoint. It has been reported that the point at which the breakpoint appears in urban environments depends on the height of the receiving antenna [79]. When the receiving antenna height (hm) is set to 2.7 m in urban districts, it is found that the actual measured breakpoint tends to be somewhat shorter than the theoretical value derived taking the reflected waves off the road surface into consideration. If one assumes for the theoretical calculation that the road surface is uniformly elevated by a certain degree, then the discrepancy between measured and theoretical values disappears. Indeed, we can interpret the presence of vehicles passing back and forth on the road as effectively raising the surface of the road. When we move the receiving antenna down to a height of hm 1.6 m, the measured results detect no breakpoint at all. This is attributed to the fact that passing vehicles frequently interrupt the propagation path, thus causing non-line-ofsight (NLOS) characteristics to appear. We obtained an average attenuation coefficient of 3.2, thus revealing a rather different property than has been obtained for propagation along road at lower frequencies. Measurements performed in the urban environment were separated into those conducted during the day and those done at night [77]. During the night there were less the one-tenth the

280

Broadband Wireless Mobile: 3G and Beyond

number of vehicles on the road as during the day and virtually no pedestrians. Using a quantitative method, we obtained distinctive breakpoint results of 320 m when measurements were done at night for hm 2.7 m and a frequency f 3.35 GHz. Comparing this measured value with the theoretical value (first Fresnel zone theory), we obtain an approximate effective road surface height h of 0.6 m. For daytime measurements under heavy traffic conditions at f 3.35 GHz it was confirmed that h 1.3 m (with a breakpoint of 170 m), and 1.4 m for the three-frequency average. It is apparent that when traffic volume is light, the effective road height is reduced and the breakpoint becomes more distant. We also obtained the breakpoint based on path loss characteristics measured during the day and at night for hm 1.6 m. At three frequencies, effective road heights ranging from 0.2 m to 0.7 m were obtained, for an average of 0.5 m. The fact that h never reached 0, even for sidewalks during periods when there were virtually no people walking about, can be attributed to the presence of trees, street lights, and cars passing on cross streets (Figures 5.8 and 5.9). NLOS path loss characteristics In this section we consider the NLOS path loss characteristics. Figure 5.10 shows typical measured path loss results as a function of distance [78], and reveals sharp jumps in losses at corners where the LOS path changes to a NLOS path. For these measurements, the transmitting base station was set up on a straight street (11 m wide) while the mobile receiver travelled down the same straight street then turned off onto two side streets, one (35 m wide) that was 64 m from the base station and the other (44 m wide) 429 m from the base station. The path loss characteristics for the NLOS portion can divided into losses that occur right at the short interval where the straight road turns the corner onto the side road (i.e., corner losses, Lc), and the section of road after that where constant attenuation coefficient a is observed. Based on measurement for this work, we found that the Lc interval, where the signal level drops precipitously, is approximately 20 m. For purposes of calculating the attenuation coefficient a, we used the entire distance

Figure 5.8

Propagation loss characteristics (hm ¼ 2.7 m).

281

Initiatives in 4Gmobile Design

Figure 5.9 Propagation loss characteristics (hm ¼ 1.6 m).

Figure 5.10

Propagation loss characteristics (hb ¼ 4.0m, hm ¼ 1.6 m, f ¼ 3.35 GHz).

including the LOS portion. At three different frequencies, we obtained the following Lc and a values for the 64-m/429-m corners, see Table 5.1. The Lc and attenuation coefficient do not reveal any dependence on frequency. It is not certain what effect the LOS distance has on Lc, but it is clear that the farther the intersection is away from the base station, the greater the attenuation coefficient. Compared to the NLOS Table 5.1

Estimated corner loss (Lc) and attenuation coefficient (a)

Frequency

Lc

3.35 GHz: 8.45 GHz: 15.75 GHz:

64 17 22 23

a m corner/429 m corner dB/16 dB dB/28 dB dB/22 dB

64 m corner/429 m corner 4.6/12 4.9/28 4.1/15

282

propagation characteristics for the UHF band (Lc rather large.

Broadband Wireless Mobile: 3G and Beyond

10 25 dB, a

4 12), Lc and a are both

Polarisation characteristics This section will cover the polarisation characteristics for microwave transmissions in an urban environment using a base station antenna installed at a low height. Measured results have already been reported for a trial that the base station (BS) antenna height, hb, was 4 m, the frequency was 8.45 GHz, and the receiving mobile station (MS) employed an antenna mounted on a moving vehicle about 3 m off the ground [121]. Comparing both of the BS and MS with vertical polarisation antenna, it was found when changing the polarisation of either of the BS or the MS that the path losses of the LOS path increased by about an average of 15 dB. Under the same conditions, the path losses for the NLOS path increased by about 9 dB. These correspond to the cross polarisation discrimination (XPD), the ratio of received power when the transmitted signal and received signal have different polarisations. In contrast to a study that obtained an XPD of 6 dB using a high base station antenna in the UHF band [108], here we used a lower base station antenna. We obtained a larger XPD and the results for the LOS path were greater than those for the NLOS path. Using the low base station antenna, the diversity effect from the polarisation tended to be rather small. Fading characteristics In mobile communications, fading often occurs even when the mobile station is not moving. This is because of the effects of vehicles and other objects present in and around the propagation path. It has been reported that these variations will differ depending on the volume of traffic even in LOS terrain [122]. Figure 5.11 shows the path loss distribution results on Rayleigh probability paper for high and low traffic volumes measured over a 200-m distance

Figure 5.11 Cumulative distributions of the path losses for different traffic conditions (hb ¼ 4 m, hm ¼ 1.3 m, f ¼ 8.45 GHz, High traffic: 1500 units/30 min.; low traffic: 600 units/30 min.).

Initiatives in 4Gmobile Design

283

between transmitter and receiver in an urban environment. As the figure makes clear, the path loss distribution approximates the Rayleigh distribution when the traffic volume is high even for LOS transmissions. When traffic is low, on the other hand, there are at least two stable components resulting from the direct wave and the wave reflected off the ground and incoherent components. As noted by Durgin et al. [25], these additive signals produce a distribution in which deeper fading occurs than in Nakagami-Rice distributions. 5.2.1.2 Propagation delay characteristics Delay profiles Delay profiles were measured by transmitting a PN-encoded signal and using a channel sounder [136]. The receiver of the sounder is stabilised by a rubidium standard signal generator. The AGC dynamic range is 50 dB, and the receiver has a dynamic range of 60 dB with a snapshot. The chip rate is 50 Mbit, and this permits measurements with a time resolution of 20 ns. Figure 5.12 shows typical measured delay profiles for an urban environment [136]. For the LOS transmission, the delay characteristics exhibit different behaviours between before the breakpoint and after the breakpoint. The results of an earlier study showing the delay spread cumulative distribution for 3.35 GHz frequency are presented in Figure 5.13 [80]. Approximating a straight line, the distribution exhibits a logarithmic normal distribution. The slope of the distribution for the domain before the breakpoint is steep, thus the delay characteristic doesn’t change much. The same tendency was observed for frequencies of 8.45 GHz and 15.75 GHz. Figure 5.14 shows the delay spread, s, as a function of distance for various receiving antenna heights in LOS transmission [81]. The delay spread increases with distance. By applying approximate straight lines to better grasp the change trends, it is apparent that the lower the hm the steeper the gradient. This can be attributed to the fact that, when the height of the antenna is approximately the same as that of vehicles and pedestrians, direct waves are obstructed and reflected rays are frequently produced by vehicles and pedestrians in the vicinity. At the same time, it was found that the delay spread as a function of distance for the NLOS transmission was several times greater than that for the LOS paths. The delay spread for urban environments was in the order of several hundred nanoseconds, about half that of suburban districts as will be discussed below. Transmission performance with RAKE reception Transmission characteristics can be estimated from measured delay profiles. In this section we shall provide an example of this [137]. RAKE reception is an effective method for dealing with multiple rays in implementing broadband multimedia transmission. RAKE receivers exploit time diversity with the same number of branches as the number of the fingers, so it is essential to know the exact number of arriving waves. The first step is to determine the number of arriving waves for a particular urban environment based on the delay profile measurement results. Here, the number of arriving waves is defined as the number of the peak powers in the delay profile within a power ratio of a [dB] from the maximum power. We confined our assessment to ratios of 3, 5, and 10 dB since RAKE reception is used for strong arriving waves. The number of arriving waves evaluation results at a frequency of 3.35 GHz are shown in Figure 5.15. One can see that the farther from

284

Broadband Wireless Mobile: 3G and Beyond

Figure 5.12 Typical urban delay profiles (hb ¼ 4.0 m, hm ¼ 1.6 m, f ¼ 3.35 GHz): (a) line of sight; (b) non line of sight.

the transmission point, the more the number of arriving wave increases, and the waves also exhibit substantial site variation. The fluctuating number of waves can be attributed to paths disappearing due to interference among multiple rays, paths disappearing as a result of nonuniform reflective surfaces of buildings, and so on. It is also clear that many sites exist where a single primary wave dominates. Focusing on a 200-meter range between transmitting and receiving antennas, at a power ratio a of 10 dB, we found that the maximum number of arriving waves was 4 at 80% of the sites. Conducting the same evaluation test at 8.45 GHz and at 15.75 GHz, no frequency-related differences could be detected. Based on these results, we will now estimate the transmission performance of 4-finger RAKE reception. Assuming a power ratio a of 10 dB, we added in-phase at most 4 powerful arriving waves, and evaluated the reception amplitude distribution. For the observation, we adopted a random phase approach that is capable of estimating propagation characteristics

Initiatives in 4Gmobile Design

Figure 5.13

285

Cumulative distribution of delay spread (hb ¼ 4.0 m, hm ¼ 2.7 m, f ¼ 3.35 GHz).

more accurately [133]. Figure 5.16 shows estimated results based on the method. Adopting RAKE reception, a 100-fold improvement in the BER was evaluated, thus the possibility of wireless transmission at rates of several megabits is shown.

5.2.2 Microwave Mobile Propagation Characteristics in Residential Environments This section gives an overview of propagation characteristics in a typical residential environment in which the average height of local houses is 8 m, the width of streets is approximately 6 m, and the height of the base station antenna is 4 m.

Figure 5.14

Delay spread/distance distribution (hb ¼ 4.0 m, f ¼ 3.35 GHz).

286

Broadband Wireless Mobile: 3G and Beyond

Figure 5.15

Arriving wave site variation.

The path loss characteristic for LOS paths is equal to an attenuation coefficient of about 2. Yet NLOS links exhibited corner losses somewhat greater than in urban environments of 30 to 40 dB, and did not always exhibit uniform attenuation with distance. Typical measurement results for a NLOS path are presented in Figure 5.17 [55]. The alternating peaks and valleys can be attributed to the waves passing along alleys and gaps between houses to reach the NLOS path and reduce the path losses. Figure 5.18 shows a typical delay profile for the propagation delay characteristic. One can see that a substantial delay occurs for the NLOS path. While the delay spread for the LOS course is no more than 200 ns, the delay spread for the NLOS path is several times greater than that [118]. Note that these values are somewhat larger than we found in earlier publications for the urban environment. This suggests that some sort of delay correction measure should be implemented for conventional UHF-band systems such as PHS. There are places where the waves passing between houses to reduce the delay are dominant, so the delay spread does not follow a monotonic ascending trend. In calculating the number of arriving waves from the delay profile, we came up with approximately the same number of waves as

Figure 5.16

BER improvement with RAKE reception.

287

Initiatives in 4Gmobile Design

Figure 5.17

Figure 5.18

Residential NLOS propagation loss characteristics.

Typical residential delay profiles: (a) LOS; (b) NLOS.

288

Broadband Wireless Mobile: 3G and Beyond

Figure 5.19

Comparison of path loss in LOS at 8.45 GHz.

when we measured the LOS path, but on the NLOS course the number of waves sometimes increased to more than 10 around cross-cutting alleys and gaps between houses [55]. Modelling method based on ray tracing The ray-tracing method based on geometrical optics theory is one well-known method of simulating propagation characteristics. The method had a number of drawbacks need for faster and more accurate simulation but a ray-tracing simulation method was recently proposed that incorporates a random phase method in a hybrid 2D 3D ray-tracing algorithm [159]. Figure 5.19 presents a comparison of actual measured results (frequency 8.45 GHz, average building height 8 m, and height of base station 4 m) and simulated results for the path loss versus distance characteristics for a LOS path in a suburban residential environment. When roadside trees are factored into the simulation, it was found that the simulated and the measured results were in very close agreement. In microwave mobile communications, moreover, the arrival angle distribution at the base station is critically important for deciding whether to make the antenna beam more narrow and for understanding the mechanism of propagation. Measured and simulated results for spatio-temporal distribution characteristics are also reported. Figure 5.20 shows a comparison of the spatio-temporal distribution characteristics done in the same way for a suburban residential environment, and one can see that the simulation results are practically identical to the actual measured results.

5.3 Adaptive Antennas 5.3.1 Introduction Adaptive antennas have long been investigated and recognised as a means of improving capacity and transmission quality of wireless systems. However, only recently has their

Initiatives in 4Gmobile Design

289

Figure 5.20 Comparison of spatio temporal distribution characteristics (for LOS at 8.45 GHz): (a) a measured azimuth delay profile; (b) an azimuth delay profile from random phase ray tracing.

performance evaluation begun to make the move from theoretical and computational to field experimental, because of the demand of higher transmission quality and rate services rather than voice-only services and also the rapid growth in the number of subscriber in recent years. Adaptive antennas may be indispensable for success of the 3G wireless systems and beyond. The practical use of antenna arrays had already started in 2G systems for cell-sectorisation, fixed (switched) multibeam antennas, beam-forming for cell design, and so on. However, the higher the transmission rate, the shorter the data symbol duration, and resolvable paths, or socalled multipath, then relatively increases without change of absolute span of the multipath. The multipath propagation channels thus have different frequency response characteristics through which transmitted signals suffer from different delay, amplitude, and phase distortions in each channel. When the bandwidth is relatively broader, received signal data at a receiver have frequency-selective faded spectrum. The channels are then called frequencyselective fading channels. The transmitted signals through the frequency-selective fading channels then overlap at the receiver and we have intersymbol interference (ISI). Furthermore, growth in the number of subscriber accelerates the increase of co-channel interference (CCI). The increase of both ISI and CCI synergically makes inefficient of the conventional uses of the array antennas. The main objective to use adaptive antenna is thus for combating the ISI and the CCI.

290

Broadband Wireless Mobile: 3G and Beyond

In this section we briefly introduce the adaptive antennas and the space-time equalisers, of which a research group in the author’s company had conducted the hardware implementations and the experimental evaluations. In the latter of this section, we present the recent investigations about CDMA adaptive array antennas and space division multiple access (SDMA) systems.

5.3.2 Algorithms Adaptive antenna algorithms that have been investigated for the use in 3G wireless system are based on minimum mean square error (MMSE) criterion using reference signal (known at receiver). A block diagram of the MMSE-criterion based adaptive antenna is shown in Figure 5.21. The adaptive antenna basically consists of the spatially-positioned M antenna elements, multipliers in which m-th received signal xm(k) is multiplied by m-th amplitude-and-phaseweight wm(k), a combiner in which all weighted signals are combined, and an adaptive weight estimator in which the variable weights can be estimated by minimising the mean square of error e(k) between the array output and the reference signals. The MMSE criterion-based adaptive antenna algorithm are described as follows:
  min E eðkÞ wT ðkÞxðkÞðkÞ2 W
 
       where w w1 w2 ; …; wM T is 1 £ M array weight vector,x k x1 k x2 k ; xM k T is 1 £ M data vector received at k-th symbol timing by the M-element array antenna,d(k) is reference signal (known at the receiver). The reference signal is usually included in transmitted data frame for synchronisation, channel estimation, and so on. Figure 5.22 shows frame structures of W-CDMA and EDGE (EDGE is a new system in convergence of GSM and IS-136). Pilot symbols in W-CDMA and sync symbols are used as reference signals for estimation of the array weights. How can we implement the MMSE-criterion-based adaptive weight control algorithm on digital signal processors? There are three widely used algorithms, as follows:

Figure 5.21

Block diagram of an MMSE criterion based adaptive antenna.

291

Initiatives in 4Gmobile Design

Figure 5.22

Frame structures: (a) W CDMA and (b) EDGE.

† Least-mean square (LMS) algorithm † Recursive least squares (RLS) algorithm † Sample matrix inversion (SMI) algorithm The SMI algorithm, which is also known as direct matrix inversion (DMI) algorithm, has recently been used for 3G systems and beyond, because the fast convergence property makes it suitable for use with high data rate transmissions [143,150]. However the complexity grows three-orders exponentially with the number of the weights (M 3). Recursive equations for the inverse of the correlation matrix thus had been used for the implementation on digital signal processors.

5.3.3 Space-time Equaliser Using Adaptive Antennas Space-time equalisers using the adaptive antennas with equalisers have also recently investigated a more powerful technique than only using the adaptive antennas [13,36 39,46,63, 104,143,144,150]. They are also called smart antennas, or intelligent antennas. Nonlinear adaptive equalisers such as decision-feedback equaliser (DFE) and maximumlikelihood sequence estimator (MLSE) had been investigated and implemented on commercial systems to compensate for the ISI. However, DFE cancels undesired delayed paths by subtracting replica from received signal and thus cannot obtain path diversity gain. MLSE, which is well-known optimum equaliser and can be implemented by Viterbi algo-

292

Broadband Wireless Mobile: 3G and Beyond

rithm (VA), but instead can obtain path diversity gain by exploiting the delayed path information. MLSE thus can have higher efficiency than DFE under the multipath-rich environment. However, the longer the span of the multipath, the more complicated the hardware implementation of the VA with the exponential behaviour in complexity as a function of the span of the ISI. Adaptive antennas on the other hand can suppress the relatively longerdelayed paths without the hardware overhead, though adaptive antennas cannot obtain path diversity gain in the same way as the DFE. The joint signal processing of the adaptive antennas and the equalisers can thus mutually compensate their drawbacks and provide higher transmission quality and capacity. Figure 5.23 shows a block diagram of the space-time equaliser [35,142 144]. The scheme proposed in [36,37] consists of a couple of adaptive antenna array processors and the branchmetric combining maximum-likelihood sequence estimator (MLSE). Here the first arrival and one-symbol-delayed path components are treated as desirable. Other longer delayed path components are suppressed as undesirable. Each array-processor combines four space-diversity branches to maximise the signal-to-interference-plus-noise ratio (SINR) of the first arrival and the one-symbol-delayed path components. One array processor combines space-diversity branches to pass the one-symbol-delayed path component into the array output with constrained first arrival path component while suppressing other longer delayed path components. Likewise, the other array combines space-diversity branches to pass the first arrival path component into the array output with constrained one-symbol-delayed path component while suppressing the other longer delayed path components. Consequently, each array processor extracts both first arrival and one-symbol-delayed path components, whose SINRs in both diversity branches are improved. Mean-square-error between the array outputs and the replicas are weighted with branch-metric-combining coefficients then combined and input to MLSE. The adjustable weights in antenna array and one-symbol-delayed tap-coefficients in the array-output-replica generator are estimated in adaptive weight controller using constrained-MMSE-criterion-based algorithm. The branch-metric-combining coefficients

Figure 5.23

Space Time equaliser.

293

Initiatives in 4Gmobile Design

can be estimated based on the quality of each diversity branch. By using the pair of array processors and the branch-metric-combining method for MLSE, a sufficient path diversity effect can be obtained when the phase differences between first arrival paths to antennas are significantly different from those on the one-symbol-delayed paths to antennas.

5.3.4 Implementation of the Space-time Equaliser The recent boom of hardware implementations of adaptive antennas and space-time equalisers may be caused by recent advances of reconfigurable hardware such as central processor units (CPUs), digital-signal-processors (DSPs), field-programmable gate arrays (FPGAs). Adaptive antennas using digital array processing is thus also called software antennas, because the digital array processing can be implemented on those programmable devises by software such as binary pattern. We had developed an experimental system using CPUs, DSPs and FPGAs, and then evaluated the performances of the adaptive antenna and the space-time equaliser [35,142,143]. Figure 5.24 shows a photograph of the experimental system and Table 5.2 describes the main specifications of the system. A lot of time and effort are still required for the development of the experimental systems for adaptive antennas and space-time equalisers, though the recent advance in the digital signal processors. We therefore developed a real-time operating system (RTOS) embedded fully programmable system for easy implementations of various space and time processing and also to carry them out simultaneously for comparison in real time. Figures 5.25 and 2.56 show the experimental results of the adaptive antenna and the spacetime equaliser. Figure 5.25 illustrates bit error rate (BER) performances under frequency-

Figure 5.24

Photograph of the experimental system.

294

Broadband Wireless Mobile: 3G and Beyond

Figure 5.25 Experimental result of the adaptive antenna (AA) and space time equaliser (AA 1 BMC MLSE) under frequency selective fading channels.

Figure 5.26

Delay time difference characteristics (3 path model).

295

Initiatives in 4Gmobile Design

selective fading channels, in which the number of arrival paths ranges from one to five and the path delays are fixed multiples of the symbol period (0, 1Ts, 2Ts, 3Ts, 4Ts). The average power is equal along all paths of each antenna. The received signal power represents the total power arriving along all paths, per antenna. Therefore, the average desired power on each path is 1/L (L: the number of arrival paths) of the total signal power received at each antenna. For the one-path model, that is, a flat fading channel, the measured BERs of both schemes are almost equal to a theoretical BER of the four-antenna maximal-ratio-combining (MRC). For the twopath model, the space-time equaliser has an improved BER because both space-and path diversity effects are obtained from signals on first and one-symbol-delayed paths. The BER of the adaptive antenna instead fell as one of its degrees of freedom is consumed in suppressing the signal on the one-symbol-delayed path. For the three-, four-, and five-path models, the pace-time equaliser has a significantly lower BER than that of the adaptive antenna. The space and path diversity effects are especially true for the five-path model because all the degrees of freedom of the adaptive antenna are used up. Figure 5.26 shows the delay time difference characteristics in the three-path model, where the delay time for the second path is set to one symbol and the delay time for the third path is varied from zero to 6Ts. The BER of the adaptive antenna increases as the difference in delay times increases. When the space-time equaliser is used, however, the BER keeps low in the range from zero to one-symbol delay because of the space and path diversity effects from signals on both first arrival and one-symbol-delayed paths. A one-symbol-delayed path does not always exist in real channels. However, the one-symbol-delayed path may be able to be produced by a delay-diversity technique [151].

5.3.5 CDMA Adaptive Array Antennas Application of adaptive array antennas is now under consideration for CDMA systems [2,5, 12,48,49,72,88 90,99,139] Using adaptive antennas at the base station, we can reduce cochannel interference, and increase the capacity of CDMA systems. Furthermore, terminals in Table 5.2

Specifications of the experimental system

Radio channel Carrier frequency RF/IF Modulation method Transmission rate Pulse shaping Array signal processing Number of antennas CPU DSP Real time operating system Viterbi equaliser (VA) FPGA Number of VA states VA path memory length

3.35 GHz/245 MHz QPSK 4.096 Mb/s Root Nyquist filter (a ¼ 0.5) 4 PowerPC 66 MHz £ 5 (Max.6) SHARC ADSP2106 (129 MFLOPS) £ 8 (Max. 40) VxWorks 3.5.1 250,000 gates 4 states 10 symbols

296

Broadband Wireless Mobile: 3G and Beyond

different angular positions can be served on the same channel with little interference if they have sufficient angular separation. Many investigations have been performed on antenna arrays of CDMA systems, including capacity evaluation, call admission control, and signal processing techniques. Most recent investigations have focused on space-time processing executed by means of antenna arrays and a RAKE receiver. In the literature [90], the spatial matched filter is performed before the despreading process and the filter outputs are despread and coherently combined by a RAKE combiner. In [99], joint space-time auxiliary-vector filtering is employed in the presence of multiple-access interference. In literatures [2,5,12,139], whole space-time processing is performed after the despreading process. The signals in different antennas are despread using the sequence of the desired terminal where the despread signal is composed of multiple delay paths. The spatial signal processing is performed for each delay path and the outputs of the spatial processors are combined by the RAKE combiner. In the spatial processors, the optimum weight vectors are given by the Wiener-Hopf solution. In [2,139], the optimal weight vectors are obtained by the normalised least mean square (LMS) algorithm with pilot symbol-assisted decision-directed coherent adaptive array diversity (PSA-CAAD). Recently, NTT DoCoMo, Japan, has carried out field experiments and laboratory experiments of PSA-CAAD with 1.990.5-MHz carrier frequency, 32-kbps information bit rate, 4.096Mcps chip rate, and Rayleigh fading environments. A recent example of technology in other literatures includes multi-user adaptive arrays with a common correlation matrix (CCM) [49], in which one common correlation matrix is used to calculate the optimal weight vectors for multiple users. Multi-user adaptive arrays with CCM can significantly decrease the computational complexity of a base station serving a number of active terminals. Another topic of CDMA systems with adaptive antennas is call admission control (CAC). With CAC, a new call is admitted if there is an available channel; otherwise the call is blocked. Since the beam pattern of an adaptive array differs terminal by terminal, a new terminal may suffer from co-channel interference even if another new terminal with a different direction does not. Therefore, the direction of the terminal must be considered in CAC. In [48], the CAC procedure is carried out by estimating new terminal’s signal-to-interferenceplus-noise ratio (SINR) at the output of adaptive array. The admission of new terminal is determined based on the estimated SINR. CDMA systems with base-station adaptive arrays are expected to achieve a capacity about 20 30% greater than that of systems with antenna diversity. More precise capacity evaluation will be required in future research.

5.3.6 SDMA (Spatial Division Multiple Access) The basic concept of spatial division multiple access (SDMA) [19,26,27,29 33,47,103,116, 147,158] is channel reuse within a cell. With the use of adaptive arrays at the base station, terminals in different angular positions can share the same time slot reducing the power of other terminals’ signals. Therefore, the SDMA system is an attractive scheme to increase the capacity of mobile communication systems. So far, the RACE TSUNAMI [147] project had field trial demonstration of both receive and transmit digital beamforming supporting SDMA systems. In addition, many literatures described beamforming methods, assignment algorithms, and power control in SDMA systems. Let us address some topics of these investigations.