Energy Conservation for IoT Devices: Concepts, Paradigms and Solutions [1st ed.] 978-981-13-7398-5;978-981-13-7399-2

Table of contents :
Front Matter ....Pages i-xii
The Rudiments of Energy Conservation and IoT (Mamta Mittal, Subhash Chandra Pandey)....Pages 1-17
Existing Enabling Technologies and Solutions for Energy Management in IoT ( Rekha, Ritu Garg)....Pages 19-47
Energy-Efficient System Design for Internet of Things (IoT) Devices (Neeta Singh, Sachin Kumar, Binod Kumar Kanaujia, Hyun Chul Choi, Kang Wook Kim)....Pages 49-74
Models and Algorithms for Energy Conservation in Internet of Things (Bhawana Rudra, Duddela Sai Prashanth)....Pages 75-110
An Energy-Efficient IoT Group-Based Architecture for Smart Cities (Lorena Parra, Javier Rocher, Sandra Sendra, Jaime Lloret)....Pages 111-127
Context-Aware Automation Based Energy Conservation Techniques for IoT Ecosystem (Monika Mangla, Rakhi Akhare, Smita Ambarkar)....Pages 129-153
Energy Conservation in IoT-Based Smart Home and Its Automation (Prithvi Pal Singh, Praveen Kumar Khosla, Mamta Mittal)....Pages 155-177
IoT Architecture for Preventive Energy Conservation of Smart Buildings (Anirudh Khanna, Shivam Arora, Anshuman Chhabra, Kartik Krishna Bhardwaj, Deepak Kumar Sharma)....Pages 179-208
Designing Energy-Efficient IoT-Based Intelligent Transport System: Need, Architecture, Characteristics, Challenges, and Applications (Kartik Krishna Bhardwaj, Anirudh Khanna, Deepak Kumar Sharma, Anshuman Chhabra)....Pages 209-233
Capacity Estimation of Electric Vehicle Aggregator for Ancillary Services to the Grid (Akhilesh Arvind Nimje, Akhilesh Baliram Panwar, Annima Gupta, Sudeep Tanwar)....Pages 235-257
Need and Design of Smart and Secure Energy-Efficient IoT-Based Healthcare Framework (Manik Sharma, Samriti, Gurvinder Singh)....Pages 259-281
Medical Information Processing Using Smartphone Under IoT Framework (Akash Gupta, Chinmay Chakraborty, Bharat Gupta)....Pages 283-308
Contributing Toward Green IoT: An Awareness-Based Approach (Suja Cherukullapurath Mana)....Pages 309-329
A New Trend to Power Up Next-Generation Internet of Things (IoT) Devices: ‘Rectenna’ (Neeta Singh, Sachin Kumar, Binod Kumar Kanaujia)....Pages 331-356

Citation preview

Studies in Systems, Decision and Control 206

Mamta Mittal Sudeep Tanwar Basant Agarwal Lalit Mohan Goyal Editors

Energy Conservation for IoT Devices Concepts, Paradigms and Solutions

Studies in Systems, Decision and Control Volume 206

Series Editor Janusz Kacprzyk, Systems Research Institute, Polish Academy of Sciences, Warsaw, Poland

The series “Studies in Systems, Decision and Control” (SSDC) covers both new developments and advances, as well as the state of the art, in the various areas of broadly perceived systems, decision making and control–quickly, up to date and with a high quality. The intent is to cover the theory, applications, and perspectives on the state of the art and future developments relevant to systems, decision making, control, complex processes and related areas, as embedded in the fields of engineering, computer science, physics, economics, social and life sciences, as well as the paradigms and methodologies behind them. The series contains monographs, textbooks, lecture notes and edited volumes in systems, decision making and control spanning the areas of Cyber-Physical Systems, Autonomous Systems, Sensor Networks, Control Systems, Energy Systems, Automotive Systems, Biological Systems, Vehicular Networking and Connected Vehicles, Aerospace Systems, Automation, Manufacturing, Smart Grids, Nonlinear Systems, Power Systems, Robotics, Social Systems, Economic Systems and other. Of particular value to both the contributors and the readership are the short publication timeframe and the world-wide distribution and exposure which enable both a wide and rapid dissemination of research output. ** Indexing: The books of this series are submitted to ISI, SCOPUS, DBLP, Ulrichs, MathSciNet, Current Mathematical Publications, Mathematical Reviews, Zentralblatt Math: MetaPress and Springerlink.

More information about this series at http://www.springer.com/series/13304

Mamta Mittal Sudeep Tanwar Basant Agarwal Lalit Mohan Goyal •

•

•

Editors

Energy Conservation for IoT Devices Concepts, Paradigms and Solutions

123

Editors Mamta Mittal Department of Computer Science and Engineering G.B. Pant Government Engineering College New Delhi, India

Sudeep Tanwar Department of Computer Science and Engineering Nirma University Ahmedabad, Gujarat, India

Basant Agarwal Swami Keshvanand Institute of Technology Management and Gramothan Jaipur, Rajasthan, India

Lalit Mohan Goyal Department of Computer Engineering J.C. Bose University, YMCA Faridabad, India

ISSN 2198-4182 ISSN 2198-4190 (electronic) Studies in Systems, Decision and Control ISBN 978-981-13-7398-5 ISBN 978-981-13-7399-2 (eBook) https://doi.org/10.1007/978-981-13-7399-2 Library of Congress Control Number: 2019935534 © Springer Nature Singapore Pte Ltd. 2019 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

The Internet of Things (IoT) has emerged significantly in recent times due to its large number of applications in variety of domains. It helps in communication among the electronic devices with the use of sensors. IoT has a significant impact on the lifestyle of people. It is estimated that IoT industry would be close to $2000 billion. In recent times, many IoT applications are becoming popular such as smart building, smart home, smart city and intelligent transportation. With the growth of the IoT devices, energy conservation is becoming quite challenging. It is very essential to understand and explore the best energy conservation technologies for better implementation of IoT devices in real time. Thus, this book focuses on energy efficiency concerns in IoT and the requirements related to Industry 4.0. The book is organized into 14 chapters. Chapter “The Rudiments of Energy Conservation and IoT” presents an introduction to the energy conservation techniques in IoT ecosystem. It presents different pragmatic energy-efficient IoT system architectures in detail. It also discusses the issues involved in the implementation of the energy conservation in IoT framework. Chapter “Existing Enabling Technologies and Solutions for Energy Management in IoT” gives an overview of IoT focusing on architectures, elements, applications and also its challenges. Authors discussed all the definitions of IoT, from various perspectives of academicians, industries and researchers. For a better understanding of IoT and its functionality, IoT architectures and its building blocks are presented in this chapter. IoT-enabled applications have been deployed in many areas. Various IoT application domains are explained in this chapter. Later in this chapter, the challenges faced by IoT are discussed and energy management is explained in detail. In the end, the future research directions are discussed. Chapter “Energy-Efficient System Design for Internet of Things (IoT) Devices” presents a brief overview of different energy-harvesting schemes along with their advantages and limitations. Authors discussed the wireless power transfer and wireless energy-harvesting scheme using rectenna technology in detail. They concluded their chapter by presenting the latest applications of IoT with their influence on human life. v

vi

Preface

Chapter “Models and Algorithms for Energy Conservation in Internet of Things” presents an overview of Big Data and how it is related to IoT and the migration of the data to the servers, load balancing among the servers, virtualization and green computing describing the importance of energy conservation along with the recent analysis of the energy consumption of the servers. Authors discussed various challenges and development of new algorithms by overcoming the limitations of the algorithms. They concluded the chapter by discussing the future challenges for the energy conservation in IoT. Chapter “An Energy-Efficient IoT Group-Based Architecture for Smart Cities” proposes an architecture that combines the data from different systems. However, in this scenario, the amount of traffic generated for each device must be considered. Therefore, authors proposed the use of data aggrupation in each antenna in order to reduce the amount of generated traffic in the fifth-generation (5G) network and thus the energy consumption. The IP of the original device and its data are grouped with IPs and the data of other devices forming a packet with all the data from one system. They tested their system with e-health data from 250 people and 3 health sensors. The results show that data aggrupation saves up to 70% of the generated traffic. Chapter “Context-Aware Automation-Based Energy Conservation Techniques for IoT Ecosystem” presents a generic framework for realizing energy-efficient smart homes, smart building and smart cities. The proposed system is based on the context-aware automation system that conserves the energy in IoT devices. Authors concluded the chapter with future directions and challenges for IoT ecosystem. Chapter “Energy Conservation in IoT-Based Smart Home and Its Automation” reviews the overall electrical network from the generation to transmission and distribution and finally to the consumer. It discusses electrical network architecture and energy monitoring points, causes of energy losses and scope for energy losses reduction. Further load shedding, utility demand energy management system using automation and the role of buildings in energy conservation have been discussed. It also discusses the design trade-offs with an emphasis on energy conservation. Authors also discussed data security model, data encryption and their practical approaches in the smart home. Chapter “IoT Architecture for Preventive Energy Conservation of Smart Buildings” scrutinizes the various architectures and frameworks currently involved with the smart building’s approach and also examines their utility, outcomes and prospects. Authors has described that how an efficient energy conservation based IoT architecture is useful in smart buildings. Chapter “Designing Energy-Efficient IoT-Based Intelligent Transport System: Need, Architecture, Characteristics, Challenges, and Applications” describes the need, scope, architecture, key technologies involved and respective power optimization prospects for intelligent transport system (ITS) providing energy-efficient solutions that can be implemented for each technical layer of ITS. Chapter “Capacity Estimation of Electric Vehicle Aggregator for Ancillary Services to the Grid” describes the capacity estimation by the aggregator and energy management strategies well in advance for bidding price and the commercial

Preface

vii

value of the electric vehicle for executing bidirectional power transfer between the vehicle and the grid. The various approaches of capacity estimation by the aggregator listed in this chapter contribute to the grid in the process of frequency regulation, thus relieving the conventional generation thereby increasing their useable life. Chapter “Need and Design of Smart and Secure Energy-Efficient IoT-Based Healthcare Framework” presents the data generation techniques used in IoT along with major applications of IoT in health care, agriculture and finance. Brief details of IoT-based publication and their trends have also been depicted. Moreover, neurological and psychological disorders have been briefly introduced in this chapter. Finally, authors presented the need and design of energy-efficient IoT-based healthcare system for psychologically and neurologically disordered patients. Chapter “Medical Information Processing Using Smartphone Under IoT Framework” presents the importance of smartphone in IoT ecosystem. Authors presented the system model of IoT e-health that gives us quality and smart way to get services which make the decision simpler and intelligent. They also presented various existing smartphone-based e-healthcare apps. Chapter “Contributing Toward Green IoT: An Awareness-Based Approach” presents various strategies for implementing green IoT by creating awareness among users. Authors discussed various ways of conducting awareness campaigns and building insight in users about the necessity of energy savings. This chapter discusses implementing smart meters to get frequent energy reading as well as indications about overuse. Smart sensors in the meters will alert the user once the usage crossed beyond the specified limit. Chapter “A New Trend to Power Up Next-Generation Internet of Things (IoT) Devices: ‘Rectenna’” demonstrates a comprehensive review of various rectenna models using different types of receiving antenna and rectifier circuit linked through a matching network. This chapter gives an overview of RF green energy systems that can be easily employed for powering next-generation IoT devices. The editors are very thankful to all the members of Springer (India) Private Limited, especially Aninda Bose for the given opportunity to edit this book. New Delhi, India Ahmedabad, India Jaipur, India Faridabad, India

Mamta Mittal Sudeep Tanwar Basant Agarwal Lalit Mohan Goyal

Contents

The Rudiments of Energy Conservation and IoT . . . . . . . . . . . . . . . . . . Mamta Mittal and Subhash Chandra Pandey Existing Enabling Technologies and Solutions for Energy Management in IoT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rekha and Ritu Garg Energy-Efficient System Design for Internet of Things (IoT) Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neeta Singh, Sachin Kumar, Binod Kumar Kanaujia, Hyun Chul Choi and Kang Wook Kim Models and Algorithms for Energy Conservation in Internet of Things . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bhawana Rudra and Duddela Sai Prashanth

1

19

49

75

An Energy-Efficient IoT Group-Based Architecture for Smart Cities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Lorena Parra, Javier Rocher, Sandra Sendra and Jaime Lloret Context-Aware Automation Based Energy Conservation Techniques for IoT Ecosystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 Monika Mangla, Rakhi Akhare and Smita Ambarkar Energy Conservation in IoT-Based Smart Home and Its Automation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Prithvi Pal Singh, Praveen Kumar Khosla and Mamta Mittal IoT Architecture for Preventive Energy Conservation of Smart Buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Anirudh Khanna, Shivam Arora, Anshuman Chhabra, Kartik Krishna Bhardwaj and Deepak Kumar Sharma

ix

x

Contents

Designing Energy-Efficient IoT-Based Intelligent Transport System: Need, Architecture, Characteristics, Challenges, and Applications . . . . . 209 Kartik Krishna Bhardwaj, Anirudh Khanna, Deepak Kumar Sharma and Anshuman Chhabra Capacity Estimation of Electric Vehicle Aggregator for Ancillary Services to the Grid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 Akhilesh Arvind Nimje, Akhilesh Baliram Panwar, Annima Gupta and Sudeep Tanwar Need and Design of Smart and Secure Energy-Efficient IoT-Based Healthcare Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 Manik Sharma, Samriti and Gurvinder Singh Medical Information Processing Using Smartphone Under IoT Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 Akash Gupta, Chinmay Chakraborty and Bharat Gupta Contributing Toward Green IoT: An Awareness-Based Approach . . . . 309 Suja Cherukullapurath Mana A New Trend to Power Up Next-Generation Internet of Things (IoT) Devices: ‘Rectenna’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Neeta Singh, Sachin Kumar and Binod Kumar Kanaujia

About the Editors

Dr. Mamta Mittal graduated with a degree in Computer Science & Engineering from Kurukshetra University in 2001 and received her Master’s degree from YMCA, Faridabad. She subsequently completed Ph.D. at Thapar University Patiala and is currently teaching at GB PANT Government Engineering College, New Delhi. She has filed two patents: for a human surveillance system, and for a wireless copter for handling and defusing explosives. She is the editor of the books “Data Intensive Computing Application for Big Data” by IOS press, and “Big Data Processing Using Spark in Cloud” by Springer. Her research interests include Data Mining, Big Data, Soft Computing and Machine learning. Dr. Sudeep Tanwar is an Associate Professor of Computer Engineering at Nirma University, Ahmedabad, India. He holds an M.Tech. in Information Technology from Guru Gobind Singh Indraprastha University, Delhi and later completed his Ph.D. in Computer Science & Engineering with a specialization in Wireless Sensor Networks. He has authored more than 50 technical research papers published in peer-reviewed international journals and conference proceedings. An Associate Editor for “Security and Privacy Journal” (Wiley), his current research interests include Routing Issues in WSN, IoT, Integration of Sensors with the Cloud, Computational Aspects of Smart Grids, and Assessment of Fog Computing in BASN. Dr. Basant Agarwal is an Associate Professor at Swami Keshvanand Institute of Technology, India. He received his M.Tech. and Ph.D. in Computer Engineering from Malaviya National Institute of Technology, Jaipur, India. He was awarded the prestigious ERCIM PostDoc Fellowship through the “Alain Bensoussan Fellowship Programme” in 2016. Having worked as a postdoctoral fellow at the NTNU, Norway, his current research interests include Artificial Intelligence, NLP, Machine Learning, and related areas.

xi

xii

About the Editors

Dr. Lalit Mohan Goyal received his M.Tech. in Information Technology from Guru Gobind Singh Indraprastha University, New Delhi, and his Ph.D. in Computer Engineering from Jamia Millia Islamia University, New Delhi. Currently, he is an Assistant Professor at the Department of Computer Engineering, J.C. Bose University, YMCA, Faridabad. He has 16 years of academic experience, and has published research papers in SCI-indexed & Scopus-indexed journals and conference proceedings.

The Rudiments of Energy Conservation and IoT Mamta Mittal and Subhash Chandra Pandey

Abstract The Internet of Things (IoT) is a new worldview which syndicates the advances and technologies of different computing domains together such as ubiquitous computing, pervasive computing, communication, and sensing technologies. In recent times, the IoT has emerged significantly as it has many applications in real time such as in smart home, smart city, smart health care, etc. In this chapter, authors have presented an introduction to the energy conservation techniques in IoT ecosystem and pragmatic energy-efficient IoT system architecture in detail. Issues involved in the implementation of the energy conservation in IoT framework as well as energy conservation approaches with its perspective are also discussed in this chapter. Keywords Internet of Things · Energy conservation · Energy efficient · Architecture · Duty cycle

1 Introduction The Internet of Things (IoT) term was coined by Kevin Ashtonin 1999. He envisioned the cascade of Internet with the physical world to upgrade the solace, security, and control of human lives. It is pertinent to define what the term “things” implicate? Perhaps, “things” can be a simple or complicated object and not necessarily be connected directly with the Internet. However, they must be connectable via a network. Broadly, IoT is the network of physical objects which entails the embedded technology which enables it to communicate and interact with the external environment. The IoT encompasses hardware, i.e., the “things”, embedded software, means to perform M. Mittal Department of Computer Science & Engineering, G.B. Pant Government Engineering College, Okhla, New Delhi, India e-mail: [email protected] S. C. Pandey (B) Department of Computer Science & Engineering, Birla Institute of Technology, Mesra, Ranchi (Patna Campus), Patna, Bihar, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. Mittal et al. (eds.), Energy Conservation for IoT Devices, Studies in Systems, Decision and Control 206, https://doi.org/10.1007/978-981-13-7399-2_1

1

2

M. Mittal and S. C. Pandey

the connectivity of the things, and information services coupled with the things. It is intuitive to ponder how does energy conservation relate to IoT? Indeed, within the purview of IoT, a huge number of devices is connected with the web and these devices are energy constrained and thus energy is a factor of paramount importance in the domain of IoT. In fact, the application of IoT is pervasive and energy should be saved and spared at different levels in IoT-based systems so that life span of different sensor nodes can be augmented. This chapter provides an overview of energy efficiency in IoT environment, as well as the means for energy conservation. The term energy efficiency covers different aspects of the domain of IoT-based systems. The most common aspects are as follows: • Energy per correctly received bit: It is the energy spent to transport one bit of information from source to the destination. • Energy per reported event: It is the energy spent to report one event. • Delay/energy trade-off: It is the notion of urgent events and speed in reporting such events. • Network Lifetime: The time in which it is able to complete its task. A sensor node utilizes its energy to do various functionalities like data collection, data communication, and data processing. Data collection is also termed as acquisition and it relies on the type of monitoring. Further, it is a fact that communication requires more energy than other processes. Furthermore, intermediate node is selected to consolidate data streams from source to sink node during the data processing phase. Nowadays, IoT is swiftly increasing its popularity in recent wireless telecommunication and alluring its importance from both academic and industry point of view [1, 2]. This technology can effectively be used in many fields like ecological monitoring, health sector, and smart cities [3–5] and in all these applications, sensors are used to collect the valuable data [6]. Indeed, IoT systems are broadly utilized as a viable transmission medium. However, it is influenced by many constraints such as battery life, memory, and transmission range [7]. In IoT, each sensor node is confined by limited battery life and transmission range. In this manner, if nodes are not able to communicate with sink node directly, they can use different companions as forwarders to route this to sink. In such case, nodes which are closer to the sink bear substantial heavy traffic loads than others. Thus, inward layer nodes exhaust their energy quicker than their peers. Thus, an immense volume of energy is wasted as other devices are far away from the sink node [8]. Indeed, advancement in sensor technology has created many new avenues pertaining to application-specific implementation of IoT. Figure 1 renders a panoramic view of the various application domains where IoT is being used for various purposes. Rest of this chapter is organized as follows: Sect. 2 will briefly elaborate the work done by different researchers in recent past pertaining to energy-efficient IoT. Section 3 delineates the energy-efficient system architecture for this pursuit. Further, Sect. 4 will render various issues related to energy conservation within the purview of IoT. Section 5 will describe different approaches of energy conservation.

The Rudiments of Energy Conservation and IoT

Industrial Control

Smart Water

3

Security and Emergency

Retails Smart Home

Smart City

Smart Environment

Logistics

Smart Animal Farming

Healthcare

Smart Agriculture

Internet of Things

Fig. 1 Application domains of IoT

Energy-efficient system design for IoT devices is discussed in Sect. 6. Finally, concluding remarks are given in Sect. 7.

2 Paradigmatic View of Energy-Efficient IoT A glimpse regarding the importance of sensors within the purview of IoT has already been given in the preceding section. Indeed, there are various kinds of sensors used in energy-efficient IoT. The paradigmatic view of sensors used in the domain of Internet of things is shown in Fig. 2. Figure 2 also delineates the interlinking among the data, process, people, things, and everyday objects. This section will briefly present the work proposed by different researchers regarding the energy-efficient IoT. Wadaa et al. [9, 10] demonstrated the energy-related problem and observed that the nodes which are nearest to the sink exhaust their energy rapidly in comparison to remote nodes. Algimantas et al. [11] proposed the basic concept of SSL protocol and proposed adaptive SSL protocol. They implemented the following security objectives in SSL protocol: confidentiality, integrity, and availability using the cryptography techniques.

4

M. Mittal and S. C. Pandey Universe of Discourse Bio Sensors Everyday Objects

NVG, Thermal Imaging Sensors RF Detectors

People

Infrared / Electro-Optical Sensors Near Field Communication

Things

Wireless Identification Sensing Platform Unique Identifier Internet of Things

Radio Frequency Identifier Spime Molecular Sensor Wireless and Sensor Actuator Sensor

Processes

Laser Audio Sensor Data

Temperature Sensor Radiation and Chemical Detector Sensing Devices

Fig. 2 The paradigm of Internet of Things

Chen et al. [12] presented comprehensive ensembles of different topology protocols and also explained different energy resource technologies like vibration energy, solar energy, and wind energy and discussed the current situation of energy supply and management techniques in the wireless sensor network. Wu et al. [13] provided an energy-efficient approach for physical layer and it is also beneficial from the aspects of deployment. They also proposed a naïve optimization principle for the purpose of energy efficiency. Suo et al. [14] discussed encryption technique, security, and shielding sensor data algorithms. They reviewed the abovementioned key technologies and adopted hop encryption protection. Mohammed et al. [15] proposed a cluster-based sleep scheduling approach. In this system, devices were deployed and clusters were formed and all devices were assumed to have the same energy. Various devices were chosen as Principal Cluster Heads (PCH) and others are made as Alternative Cluster Heads (ACH). These alternative heads are used to provide fault tolerance in case PCH devices fails. In each cluster, few devices are active and used in network coverage whereas other devices remain inactive/sleep state. Thus, the energy consumption was lower compared to other methods.

The Rudiments of Energy Conservation and IoT

5

Abbas et al. [16] performed an exhaustive study on energy preserving issues and solutions. In [16], authors have handled various issues of IoT such as duty cycle, congestion control, sleep and wake-up time and determination of heterogeneous radio interfaces. Future research directions to preserve energy were also proposed in [16]. Keshavarzian et al. [17] studied the peculiar behavior of nodes termed as idle listening. The term idle listening is defined as the consumption of extensive part of power on checking the encompassing condition as opposed to communicating with other nodes. In [18–20], the duty cycle operation-based methodologies have been proposed. In this, devices are periodically flip-flopping between sleeping and active mode and thus strive to reduce the energy consumption. This protocol is widely acceptable in IoT network designing [18–20]. In [21–23], it has been mentioned that clustering structure is more effective way to deal with energy-related issues and it also drags out network lifetime issues.

3 Pragmatic Energy-Efficient IoT System Architecture From the viewpoint of energy conservation, IoT system architecture entails four stages as enumerated below [24]. • • • •

Stage 1. Sensors/actuators Stage 2. The Internet gateways Stage 3. Edge IT Stage 4. The data center and cloud

Stage one is associated with the sensors and actuators. The role of sensors is to collect data for the purposes of analysis. Further, actuators perform substantial role in the alteration of physical conditions during the process of data recording. It is further followed by the conversion of analog data into digital streams with the help of data acquisition system (DAS). The function of Internet gateway is to route the digitized data over Wi-Fi, wired LANs, or the Internet. Further, it is of the paramount importance to incorporate the Edge IT in order to prevent the inundation of data center resources. The Edge ITs are located in remote areas and edges considerably reduce the burden on core IT infrastructure. Edge IT is also indispensable to save the network bandwidth; support the security concerns; storage issues; and delays the processing of data. Finally, the last stage can be utilized for data which requires in-depth processing and considerably more computational power. Moreover, cloud and data centers further execute a more in-depth analysis. Indeed, from the perspective of energy efficiency, a universally accepted architecture design is yet to be worked out. However, few recognizable designs include [25]:

6

• • • •

M. Mittal and S. C. Pandey

Three-Layer Architectures Five-Layer Architectures Cloud- and Fog-Based Architectures Social IoT

Three-layer architecture is the most basic form of architecture and it comprises of three layers. Namely: application layer, network layer, and perception layer. The role of the application layer is to provide services to the users and the functioning of the network layer is to provide the connectivity to all part of the network. The perception layer is a physical layer and it incorporates the physical components such as the sensors, actuators, etc. Five-layer architecture portraits a better and wider view of IoT architecture and it is given in Fig. 3. In this architecture, the business layer performs an entire business end of the whole IoT system. It mostly includes prototypes that find profit and business models. Further, transport layer acts as a permeable interface connecting both processing and perception layers. The next layer is the processing layer and the task of this layer is to process, analyze, and stores the chunk of data. In peculiar circumstances such as when data utilization and processing is an arduous task owing to gigantic volume of data, organizations often require a separate data center platform. Cloud-centric architectures can play substantial role in such a position. In cloud-centric architectures, cloud is considered at the uppermost level and it also possesses the central role of processing because of its scalability in such operation [26]. Further, fog computing is analogous to edge IT and its function is to lower down the process strain on the center platform. Fog architecture also entails the preprocessing of data at sensors and Internet gateway level. Social IoT is a new paradigm of system’s architecture and it is based on real-life socialization among humans. Modeling is done according to human social networks. A different layer of cloud and fog-based architecture, and social IoT is also shown in Fig. 3. It also delineates the stages required in the architectural design of IoT and a layer- based dichotomy of different architectures.

4 Issues of Energy Conservation in IoT It is intuitive to think that in near future IoT will change the manner we deal with our life. It is explicit that IoT is creeping inside gradually in different walks of our life including homes and workplaces. Indeed, IoT has left its permanent impression in the realms of consumerism and business and perhaps mitigation in energy consumption is an important aspect and it is worthy to analyze the different issues which render energy conservation in IoT. It is empirical fact that energy conservation is an arduous task particularly when operational costs are going to increase whereas resources shrink. IoT is the huge system of web associated things and it empowers the client to gather and investigate information using associated devices. Further, the information and explicitly the

The Rudiments of Energy Conservation and IoT

7

Stages of Architecture Stage 1: Sensors / Actuators

Stage 2: The Internet Gateway

Stage 3: Edge IT

Stage 4: The Data Centre and Cloud

Types IoT System Architecture

Three-Layer Architecture

Five-Layer Architecture

Cloud and Fog based Architecture

Application Layer

Business Layer

Transport Layer

Network Layer

Application Layer

Security Layer

Perception Layer

Processing Layer

Storage Layer

Transport Layer

Preprocessing Layer

Perception Layer

Monitoring Layer

Physical Layer

Fig. 3 Architectural details of IoT systems

8

M. Mittal and S. C. Pandey

associated productive knowledge can be utilized to enable entrepreneurs to settle on cultivated choices about everything from energy acquisition to its day-to-day practice. The final product should utilize the energy in an exceptionally effective manner so that utility cost can be reduced. Different issues pertaining to energy conservation can be categorized as: • • • •

Day-to-day energy engagement Smart analysis and planning Proactive and preventative maintenance Employee involvement and satisfaction.

It is a common phenomenon that we often uselessly waste the energy inside the home. It is also true that the innocuous actions of wasting of energy often become victims of oversight. Often, lights are left on and indoor regulators are left to our most extreme comfort temperature while no one is inside the home. Perhaps, business is also not protected from such negligence and after some time, absent-mindedness, and wasteful energy practices will impose gigantic levy. Indeed, smart innovation can respond to empty rooms, turning off our lights and air conditioners and saves the energy. It is thus obvious that this technology possesses the efficacy to empower us to utilize and control the energy usage optimally everywhere. In addition to reacting and meeting the energy needs of a facility, IoT devices collect data that can have a significant and long-lasting impact on future usage. Perhaps, this accumulated information renders the business owners to identify the potential issues and subsequently enable them to make smart decision. Additionally, the data collected can also identify issues that may have previously gone unnoticed. For example, a hike in energy use in a specific zone may indicate potential problems with heating or cooling systems or faulty or poorly performing equipment—both of which can attribute to unnecessarily high costs. Indeed, advanced sensors are required in IoT devices so that it can alert the business owners about the potential issues well in advance before they turn out to be big issues. We can consider the case of a cracked pipe which can cause huge loss of costly chemicals. Sensors can tackle such issues immediately before real damage happens. A similar type of things can exist inside the hardware devices (machinery), where sensors can recognize changes in sound and new vibrations which may go totally unnoticed by the bare eye or ear. These issues can increase energy consumption level until the point that the issue is notable and resolved. Finally, and maybe above all, it is beneficial from the human perspective. We will most likely be unable to control the manner in which workers/employees’ approach or deal with our energy preservation activities, so, with the help of new, innovative strategies and connected facilities owners can monitor their performance and encourage them. Business owners can provide incentive and reward the departments who can decrease the energy consumptions. It is thus obvious that the owner can control the environment and provides the opportunity for maximum human comfort without unnecessary cost inflation. Further, a comfortable, contended, and happy employee can help more in productivity and undoubtedly upgradation in existing facilities as well as redesigning of the office by incorporating IoT innovations can

The Rudiments of Energy Conservation and IoT

9

Collision Listening

Traffic Fluctuation

Energy Efficiency in IoT Over-Hearing Listening

Idle Listening Protocol Overhead Reduction Fig. 4 Major issues related to energy efficiency of IoT

help in this pursuit. It is the empirical fact that smart spaces lead to smart decisions and those decisions will change everything from budget to organization culture. Furthermore, major issues pertaining to energy efficiency of IoT is depicted in Fig. 4 and also enumerated below. • • • • •

Idle Listening Collision Overhearing Reduction of protocol overhead Traffic Fluctuation

Energy utilization is major while the node is in active mode. It is even true when node is sitting idle and waiting for its turn to transmit the data. In the idle state, the node is only monitoring its surroundings rather than receiving or sending the packets. In order to overcome this issue, provision is there to turn back the sensor nodes from sleeping mode to active mode after a predefined interval of time or after processing of a wake-up signal. Collision occurs when the nodes get different data packets in the meantime. Further, because of the collision, all the received data become pointless. In this case, retransmission has to be carried out which will cause excess energy consumption. Collision increases latency as well and these transactions could take considerable amount of energy. During the data transformation, interference may occur with the neighboring node which is termed as overhearing. It has been observed that the nodes within reach possess this particular problem and leads to burn up energy resources. The protocol header information also exhausts the energy resources. That’s why it is recommendable to use some methods to reduce the overhead such as adaptive transmission periods, cross-layering approaches, and optimized flooding. Fluctuations in the network traffic may cause high delay or congestion. Further, if there is maximum traffic on the network provided the network is functioning with its utmost efficacy then congestion will raise up to a substantial apex level.

10

M. Mittal and S. C. Pandey

Fig. 5 Diagrammatic representation of energy conservation approaches for IoT devices

5 Energy Conservation Approaches for IoT Devices and Its Perspectives In literature, different approaches have been proposed for the purpose of energy conservation. Each approach has its own merits and demerits. Selection of a specific approach for energy conservation is a critical issue as it depends on many factors. This section preludes some approaches of paramount importance used for the purpose of energy conservation in IoT paradigm. Figure 5 shows a diagrammatic representation of energy conservation approaches for IoT devices.

5.1 Node Activity Management There are two parts in node activity. Namely; sleep scheduling and on-demand node activity. In sleep scheduling, time of sleeping mode as well as the wake-up timing is set a priori. This saves energy in idle time spans. On-demand node activity is not scheduled; rather node is in active state with some functionality. When a wake-up signal is communicated/broadcasted, the neighboring nodes within the area change their state to active mode and subsequently, the data transmission takes place. In addition to it, an activity scheduling scheme for sensing the coverage has also anticipated [27]. It is worthy to mention that this is done periodically and every time a random timeout

The Rudiments of Energy Conservation and IoT

11

is opted by the node and it gets tuned with the messages coming from neighboring nodes before its termination takes place. This message comprises of the decision whether the node would be active or not and thus small energy is required to pass such message which in turn reduces the energy consumption at the node level.

5.2 Data Aggregation and Transmission Process It is explicit that data transmission costs more than its processing. Therefore, it is prudential to aggregate the data inside the cluster. In this process, data which is coming from different sources is consolidated in one packet. This helps to reduce redundancy and minimize the number of transmissions. Wireless transmission of data utilizes the aggregate energy profusely. Thus, if power control is incorporated in the transmission process, high energy saving will occur [28]. Further, in order to accomplish high energy efficiency, transmission control and circuit power should be carefully adjusted. Over and above, optimization algorithms can also be used for this purpose [29].

5.3 Media Access Control (MAC) Protocol Energy is very important factor while dealing with IoT devices. Better design of MAC protocol is one of the ways to use energy efficiently. Indeed, energy efficiency is the most important attribute of MAC Protocol. MAC protocol describes the rules to transmit the frame. If there are many nodes, then it synchronizes the channel access. A popular MAC standard is described by the IEEE in 2003 and further revised in 2006. There are two modes of operation in MAC standard as defined in IEEE 802.15.4. These are (1) Non-beacon-enabled and (2) Beacon enabled. The former is always in a awaken state to receive a frame. The latter defines superframes in which nodes are awaked just for a small portion of a super frame. This causes increase in energy utilization. The basic idea of duty cycle is to reduce the pointless action by setting the node in the sleep state. In this, periodic wake-up scheme is implemented which wakes the node up occasionally to transmit or receive packets. If there is no activity, the node comes back to sleep state. There are different low duty cycle protocols which are delegated into synchronous and asynchronous fashion. The idea of synchronous is associated with data exchanges. There are two fundamental ideas in asynchronous mode one is transmitter initiated and other is receiver initiated. In the former case, a node frequently directs request packets until reaches the destination. In the latter approach, a node sends packets to notify the neighboring nodes about the inclination to get the packets. To accomplish low power operation, other MAC protocols often used in IoT are MQTT, XMPP, DDS, and AMQP. It has been proposed by the researchers that energy could be efficiently handled by proper design of MAC protocol.

12

M. Mittal and S. C. Pandey

5.4 Security Management Energy is an important factor to consider security measures for the nodes. But security systems are not intended from resource point of view. Thus, there exist many challenges like how the encryption algorithms work speedily with little energy consumption. The existing techniques require powerful equipment and thus it is of utmost importance to confine the energy utilization and thereby extend the battery life. Indeed, security measure is required in each layer of IoT and security measures render significant impact on energy consumption in order to perform the encryption and decryption functions. In the perceptual layer, confirmation is important to avoid illegal access to the node. Further, to ensure the secrecy of information, data encryption is necessary and stronger security measures consume more energy. In the network layer, it is difficult to apply the existing communication security mechanisms which consume more energy because confidentiality and integrity are important in this layer. Furthermore, there are two aspects in application layer, i.e., confirmation and key understanding across the heterogeneous network. In fact, the systems are not designed for resource-restricted devices. Therefore, lightweight cryptographic algorithms are needed to be implemented in this layer.

5.5 Topology Management The role of topology control is to reduce the node power consumption and thus to extend the network lifetime. The types are as follows: • • • •

Graph-Based Topology control Relative neighborhood Graph Gabriel Graph Localized Minimum Spanning Tree

Graph-based topology control is done locally provided the data about separations among the sensors and their relative position is accessible. Relative neighborhood graph is a straight-line chart to connect two points. The triangulation of a point is maximal of non-intersecting line segments with vertices in the point. Further, in Gabriel graph every sensor in the network is aware of the neighboring sensors as well as their locations. This graph determines the logical neighbor of the sensor by calculating the closed discs of diameters. In addition, localized minimum spanning tree processes a power diminished network topology by developing a minimum spanning tree over the network in a completely dispersed way. Energy consumed is less in this topology as compared to the original network.

The Rudiments of Energy Conservation and IoT

13

5.6 Routing Routing is the act of transferring information across network from source to destination. It occurs in the network layer and makes the decision of which route is to be used. It is categorized in different groups such as flat-based, hierarchical-based, and location-based routing. In the first method, an equivalent job or functionality is assigned to each node. In the second method, nodes will assume distinctive roles. Further, in the third method positions of the nodes is a decisive factor for the purpose of routing the data. The characteristic of routing protocols can be arranged into three heads. Namely: (1) Proactive (2) Reactive, and (3) Hybrid. The first protocol collects routing data proactively, endeavoring to have an impression of the whole network’s topology consistently. A reactive protocol searches the routes on-request and when a transmission begins the route discovery process is triggered. Hybrid is the combination of both. Packet sending might be either hop by hop or through source routing. In hop-by-hop router, only a small part of each router is used whereas in other methods, complete path of router is used to transfer the data. Hop count is the most common metric and this metric may suggest the routing path selection for the conservation of energy [30].

6 Energy-Efficient System Design for IoT Devices It is anticipated that by the coming decade, there will be more than 50 billion smart substances associated with IoT. These smart substances will create interfacing between the external world and computing framework. Perhaps, this interfacing will alter all parts of our everyday lives and reform various application areas such as medicinal services, energy protection, transportation, and so on. Designing aspects of IoT devices is a challenging task because IoT devices comprise of different components like sensors, actuators, and intuitive devices. These devices are connected together through network and collecting data using sensors. Most significant design principles for the IoT devices are shown in Fig. 6 and also enumerated below: • • • • • • •

Focus on value. Take a holistic view. Put safety first. Consider the context. Build a strong brand. Prototype early and often. Use data responsibly.

In the realm of IoT, designing of services according to user’s need is a critical issue and it seems that early researchers are anxious to experiment with new innovation while numerous others are hesitant to bring new innovation into utilization. Perhaps,

14

M. Mittal and S. C. Pandey Safety

Focus on value

Physical World

Holistic View

Energy-efficient system design forIoTdevices

Trust

Context Prototype

Data Usability

Fig. 6 Panoramic view of significant design principles for the IoT devices

it is due to the fact that they were not feeling sure with it. Indeed, it is very important to focus on the real problems faced by the end user and how our IoT solution will meet the user’s need so that it can be widely accepted. Further, it is also of the utmost importance to comprehend the obstacles coming forth in the adoption of the new innovation and how IoT solution explicitly solves it. It is intuitive to consider the highlights that may be profitable and profoundly applicable and that might be uninteresting for most of the users and vice versa. Therefore, it is indispensable to think about the incorporation of the features and their order in the designing process. IoT services normally comprise of various devices with physical and digital touch points. These devices can perform different functions and having different capacities. Moreover, they may be work in cooperation with distinctive service providers. Therefore, it is not sufficient to plan one of the touch points well; rather we have to investigate the entire framework. The designer of IoT devices must consider various issues like the job of every device and benefit, and the intangible model of how user comprehends and observe the framework so that the entire framework can work consistently together to render a synergize outcome. As IoT solutions deal with the real-life problems, its consequences can be dangerous if something turns out badly. Moreover, the users are also varying according to the IoT solutions and utilizations of new innovation. Therefore, trust formation is the key issue in this pursuit. Perhaps, trust formation takes place gradually and lost effectively so it is worthy to ensure that everyone associated with the products or services must be benefited which in turn will bring more trust instead to breach it. Trust formation requires understanding of the conceivable error possibilities identified with setting of utilization including hardware, software, and network. In addition, interaction

The Rudiments of Energy Conservation and IoT

15

with the users and endeavors to avoid users problem beforehand also plays pivotal role in trust building. Data security and privacy are also the key components of the design phase. It is extremely imperative for the users to feel that their private information is protected. Users must feel that their home, working condition, and everyday items cannot be hacked and their friends and family are not put in danger. Lastly, quality assurance is basic and it must be reflected in software testing as well as in the end framework. In IoT devices command provide the digital interfacing with the real world and adverse command can cause severe impact on the real world and unlike digital commands, it is not possible to undo such adverse command. The success of IoT devices is dependent on both the hardware and software. So, the life span of both the hardware and software should be adjusted properly while designing the IoT devices. Further, IoT devices are very difficult to upgrade because once the connected devices are set it is not so natural to replace them with a more current adaptation and even the software inside the connected devices might be difficult to refresh because of security and protection reasons. Because of these aspects and to avoid expensive hardware interaction, it is critical to get the appropriate solution from the commencement of implementation. Thus, it implies that prototyping as well as rapid interaction of both the hardware and the entire system is must during the early phases while designing the devices. Perhaps, new and progressively inventive methods for prototyping and imitating the solutions are required. IoT devices can undoubtedly produce abundant amounts of data. Therefore, it is fundamental for the designer to comprehend the potential outcomes of data science. Data science gives a great deal of chances to minimize the user’s grudge by reducing the utilization of time, energy, and stress. Data science also helps to automate the decision, to filter important signals from noise as well as to infer even from the deficient/lacking data input. Indeed, understanding how data can be accessible and how it may be utilized to help the user is a key component in designing efficacious IoT Solutions.

7 Conclusions In this chapter, authors have presented the outline about the importance of sensors within the purview of IoT and also preludes different paradigms of energy-efficient IoT systems. Further, different stages which IoT system architecture entails from the viewpoint of energy conservation are discussed. Furthermore; different issues pertaining to energy conservation are categorized and analyzed. Moreover, energy conservation approaches for IoT devices are also incorporated in this chapter. Designing aspects of IoT devices is a challenging task. However, this chapter describes the most significant design principles for the IoT devices in a lucid manner.

16

M. Mittal and S. C. Pandey

References 1. Zhang, D., Zhou, Z., Mumtaz, S., Rodriguez, J., Sato, T.: One integrated energy efficiency proposal for 5G IoT communications. IEEE Internet Things J. 3(6), 1346–1354 (2016). https:// doi.org/10.1109/JIOT.2016.2599852 2. Gubbi, J., Buyya, R., Marusic, S., Palaniswami, M.: Internet of things (IoT): a vision, architectural elements, and future directions. Future Gener. Comput. Syst. 29(7), 1645–1660 (2013). https://doi.org/10.1016/j.future.2013.01.010 3. Islam, S.R., Kwak, D., Kabir, M.H., Hossain, M., Kwak, K.-S.: The internet of things for health care: a comprehensive survey. IEEE Access 3, 678–708 (2015). https://doi.org/10.1109/ ACCESS.2015.2437951 4. Solanas, A., Patsakis, C., Conti, M., Vlachos, I.S., Ramos, V., Falcone, F., Postolache, O., Perezmartinez, P.A., Pietro, R.D., Perrea, D.N., Martinez-Balleste, A.: Smart health: a context-aware health paradigm within smart cities. IEEE Commun. Mag. 52(8), 74–81 (2014). https://doi. org/10.1109/MCOM.2014.6871673 5. Zanella, A., Bui, N., Castellani, A., Vangelista, L., Zorzi, M.: Internet of things for smart cities. IEEE Internet Things J. 1(1), 22–32 (2014). https://doi.org/10.1109/JIOT.2014.2306328 6. Li, Q., Ding, D., Conti, M.: Brain-computer interface applications: security and privacy challenges. In: 2015 IEEE Conference on Communications and Network Security, CNS, pp. 663–666 (2015). https://doi.org/10.1109/CNS.2015.7346884 7. Zhu, Q., Wang, R., Chen, Q., Liu, Y., Qin, W.: IoT gateway: bridging wireless sensor networks into internet of things. In: 2010 IEEE/IFIP International Conference on Embedded and Ubiquitous Computing, pp. 347–352 (2010). https://doi.org/10.1109/EUC.2010.58 8. Liu, A.-F., Wu, X.-Y., Chen, Z.-G., Gui, W.-H.: Research on the energy hole problem based on unequal cluster-radius for wireless sensor networks. Comput. Commun. 33(3), 302–321 (2010). https://doi.org/10.1016/j.comcom.2009.09.008 9. Wadaa, A., Olariu, S., Wilson, L., Eltoweissy, M., Jones, K.: Training a wireless sensor network. Mob. Netw. Appl. 10(1–2), 151–168 (2005). http://dl.acm.org/citation.cfm?id=1046430. 1046442 10. Wu, X., Chen, G., Das, S.K.: Avoiding energy holes in wireless sensor networks with nonuniform node distribution. IEEE Trans. Parallel Distrib. Syst. 19(5), 710–720 (2008). https://doi. org/10.1109/TPDS.2007.70770 11. Venckauskas, A., Jusas, N., Kazanavicius, E., Stuikys, V.: An energy efficient protocol for the internet of things. J. Electr. Eng. 66(1), 47–52 (2015) 12. Chen, F., Guo, L., Chen, C.: A survey on energy management in the wireless sensor networks. Elsevier 3, 60–66 (2012) 13. Wu, G., Yang, C., Li, S., Li, G.Y.: Recent advances in energy-efficient networks and their applications in 5G systems. IEEE Wirel. Commun. 22(2), 145–151 (2015) 14. Suo, H., Wan, J.: Security in the internet of things: a review. IEEE 3, 648–651 (2012) 15. Al-kahtani, M.S.: Efficient cluster-based sleep scheduling for M2M communication network. Springer 40(8), 2361–2373 (2015) 16. Abbas, Z., Yoon, W.: A survey on energy conserving mechanism for the internet of things: wireless networking aspects. Sensors 15, 24818–24847 (2015) 17. Keshavarzian, A., Lee, H., Venkatraman, L.: Wakeup scheduling in wireless sensor networks. In: Proceedings of the 7th ACM International Symposium on Mobile Ad Hoc Networking and Computing, MobiHoc’06, ACM, pp. 322–333 (2006). https://doi.org/10.1145/1132905. 1132941 18. Palattella, M.R., Accettura, N., Grieco, L.A., Boggia, G., Dohler, M., Engel, T.: On optimal scheduling in duty-cycled industrial iot applications using IEEE802.15.4e TSCH. IEEE Sens. J. 13(10), 3655–3666 (2013). https://doi.org/10.1109/JSEN.2013.2266417 19. Lin, P., Qiao, C., Wang, X.: Medium access control with a dynamic duty cycle for sensor networks. In: 2004 IEEE Wireless Communications and Networking Conference, IEEE Cat. No. 04TH8733, vol. 3, pp. 1534–1539 (2004). https://doi.org/10.1109/WCNC.2004.1311671

The Rudiments of Energy Conservation and IoT

17

20. Medidi, M., Zhou, Y.: Extending lifetime with differential duty cycles in wireless sensor networks. In: IEEE GLOBECOM 2007 - IEEE Global Telecommunications Conference, pp. 1033–1037 (2007). https://doi.org/10.1109/GLOCOM.2007.199 21. Soro, S., Heinzelman, W.B.: Prolonging the lifetime of wireless sensor networks via unequal clustering. In: 19th IEEE International Parallel and Distributed Processing Symposium, pp. 8–15 (2005). https://doi.org/10.1109/IPDPS.2005.365 22. Yang, M., Wang, S., Abdelal, A., Jiang, Y., Kim, Y.: An improved multi-layered architecture and its rotational scheme for large-scale wireless sensor networks. In: 2007 4th IEEE Consumer Communications and Networking Conference, pp. 855–859 (2007). https://doi.org/10.1109/ CCNC.2007.173 23. Mhatre, V., Rosenberg, C.: Design guidelines for wireless sensor networks: communication, clustering and aggregation. Ad Hoc Netw. 2(1), 45–63 (2004). https://doi.org/10.1016/S15708705(03)00047-7 24. Fuller, J.: How to design an IoT-ready infrastructure: the 4-stage architecture (2018). https:// techbeacon.com/4-stages-iot-architecture 25. Sethi, P., Sarangi, R.S.: Internet of things: architectures, protocols, and applications (2017). https://www.hindawi.com/journals/jece/2017/9324035/cta/ 26. Benazzouz, Y., Munilla, C., Gnalp, O., Gallissot, M., Grgen, L.: Sharing user IoT devices in the cloud. In: 2014 IEEE World Forum on Internet of Things (WF-IoT), pp. 373 (2014). https:// doi.org/10.1109/WF-IoT.2014.6803193 27. Katkar, P.S., Ghorpade, V.R.: A survey on energy efficient routing protocol for wireless sensor networks. Int. J. Comput. Sci. Inf. Technol. (IJCSIT) 6(1), 81–83 (2015) 28. Zheng, W., Rahman, K.A., Tepe, K.E.: Reliable and energy efficient cooperative relaying scheme (REECR) in wireless sensor networks, IEEE, pp. 400–405 (2013) 29. Kumar, C., Paulus, R.: A prospective towards M2M communication. J. Converg. Inf. Technol. (JCIT) 9, 102–114 (2014) 30. Talwar, M.: Routing techniques and protocols for internet of things: a survey. Proc. NCRIET Indian J. Sci. Res. 12, 417–423 (2015)

Existing Enabling Technologies and Solutions for Energy Management in IoT Rekha and Ritu Garg

Abstract Internet of Things which is an emerging technology extends the internet deep into the physical environment. This chapter gives an overview of IoT focusing on architecture, elements, applications and also its challenges. Though IoT is defined in many ways, no single definition is worldwide agreed. The basic concept of IoT is that everyday objects which are capable of identifying, sensing and processing can communicate with one another and services over the Internet to accomplish some useful objective using networking concepts. Many definitions of IoT, from various perspective of academician, industries and researchers, are also discussed. For better understanding of IoT and its functionality, IoT architectures and its building blocks are presented in this chapter. IoT enabled applications have been deployed in many areas. Various IoT application domains are explained in this chapter. Later in this chapter, challenges faced by IoT are discussed and energy management is explained in detail. At the end, the future research directions are discussed. This chapter provides an insight on basics of IoT. After going through this, reader will be able to understand the IoT architecture, its elements and applications. It also gives the review of issues in IoT and energy management. Keywords Internet of Things · Architecture · Components · Energy management · Application · Energy conservation · Energy harvesting

1 Introduction In 1998, Kevin Ashton introduced a new paradigm in the area of wireless communication and networking, known as “Internet of Things (IoT)” [1]. IoT enables the system to attain automation and analysis. It exploits existing and emerging technology for sensing, networking, and robotics. IoT utilizes modern advancements in software, Rekha (B) · R. Garg Computer Engineering Department, NIT Kurukshetra, Kurukshetra, India e-mail: [email protected] R. Garg e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. Mittal et al. (eds.), Energy Conservation for IoT Devices, Studies in Systems, Decision and Control 206, https://doi.org/10.1007/978-981-13-7399-2_2

19

20

Rekha and R. Garg

falling hardware prices and recent attitudes towards technology. Its advanced features have impact on social, politics and economics due to considerable changes in delivery of services, goods and products. A closer look at the phrase Internet of Things (IoT) provides a vision of two main pillars of IoT: Internet and Things which need more explanation. IoT is making world more connected by connecting almost all objects to Internet. These objects fall into the category of “Things”. This category includes environmental elements, actuators, human beings, sensors, Radio-Frequency IDentification (RFID) tags, etc. [2, 3]. Many definitions of IoT, from various perspectives of academician, industries and researchers, are given. Some definition given by researchers are: – “A universal framework, based on existing and emerging information and communication technologies, for information sharing utilizing the advanced services by connecting the physical and virtual things” [4]. – “IoT is 3 A concept which means any media can be connected to anywhere anytime, resulting into sustained ratio between radio and man around 1:1” [5]. – “A self-configuring dynamic universal network framework based on standard and interoperable communication protocols where identifiable physical and virtual ‘Things’ having physical attributes use intelligent interfaces and are seamlessly integrated into the information network” [6]. – “Identifiable things having virtual personalities functioning in smart spaces utilize intelligent interfaces for connection and communication within environmental, social, and user contexts” [7]. – “IoT is an interaction between the digital and physical worlds using an abundance of actuators and sensors” [8]. – “IoT can be defined as embedding computing and networking capabilities in various types of object” [9]. The phrase Internet-of-Things is a hypernym term covering different aspects of extending Web and Internet into physical world, by enabling worldwide distribution of devices capable of identification, sensing or actuation. So in simple words, “Internet of Things” can be defined as connection of any devices to the internet for communication and data analysis via embedding the sensor, software and actuator. IoT is making world widely open, offering virtually endless scopes and connections at office, house, etc. Figure 1 depicts an example of scenario of Internet of Things with different application domains of IoT. According to a report provided by Gartner, by 2020 25 billion devices will be tethered to Internet [10]. These connected devices will provide the facility to analyse, manage the data and help in making smart decisions autonomously. We can see that IoT is widely used in many sectors, such as e-education, e-governance, agriculture, smart domestics, retail automation, transportation, industrial manufacturing, business management, smart city, smart health, assisted living and logistics etc. In 52 countries, there are 1987 IoT-funded companies working in 20 categories [11]. There exist several definitions of IoT, some of which are discussed above. Our rest of chapter is organized as follows. Section 2 gives insight into various IoT architectures. Section 3 discusses several building blocks of IoT. IoT has been deployed in

Existing Enabling Technologies …

21

Fig. 1 A Simple Scenario of Internet of Things with the end users and different application domains

various sectors. Several IoT application domains are explained in Sect. 4. Followed by this, in Sect. 5, various challenges faced by IoT are introduced. Out of these challenges, energy management is explained in detail in Sect. 6. The future research directions are given in Sect. 7.

2 Architectures of IoT The IoT connects trillions of devices/objects which can produce lots of traffic and to manage this huge amount of data storage will be needed. Several architectures have been proposed by the researchers because there is no single agreement for IoT architecture, which is agreed worldwide. In this section, several architectures of IoT are discussed.

22

Rekha and R. Garg

2.1 Three-Layer Architecture Figure 2 depicts the most basic three-layer IoT architecture. It was proposed in the early state in the area of research. The three layers in this architecture are: the perception, network and application layers [12, 13]. The description of these layers is given below. 1. Perception Layer: This layer is similar to physical layer in OSI model and also known as “Device Layer”. The aim of this layer is identification, tracking, information acquisition, and processing from the environment. For this purpose, this layer is composed of various types of physical devices and sensors such as Zigbee, Infrared, RFID, QR code, etc. These devices systematically monitor physical world and collect the data. Based on the sensor type, gathered data can be dust amount in air, temperature, location, pH level, wind speed, humidity, vibration, etc. Information collected is transmitted to network layer for secure communication. Perception layer forms the basis for IoT system via enabling technologies like short-range radio, signal detection, and so on. 2. Network Layer: This layer provides the facility to connect with other smart objects. This layer is also responsible for routing and secure data transmission. Network layer guarantees for secure information transmission to destination through Wi-Fi, 3G, infrared, Bluetooth, UMTS, Zigbee, RFID, Satellite, WiMAX, etc. depending on sensor device used. Features of this layer are also used to transmit and process the sensor information. IEEE 802.15.4, 6LoWPAN, Zigbee, Z-Wave, etc., are some protocols for secure and reliable communication in IoT. 3. Application Layer: AT the top of stack, application layer provides the application management. The information received from the lower layer is managed by respective management system and services specific to application are delivered to the user. There are numerous IoT application areas like: health monitoring, manufacturing, smart city, smart home, transportation, public safety, etc.

Fig. 2 Three-layer architecture

Existing Enabling Technologies …

23

2.2 Four Layer Architecture SoA has been already exploited successfully in different research domains such as cloud computing, WSNs, etc. In many systems, three-layer architecture of IoT has been realized and designed. Though, the multi-layer IoT architecture is simple, but operations and functions performed by the application and network layers are both complex and diverse. For example, routing and data transmitting are not only tasks performed by network layer, but it also provides data services such as data aggregation, computing, etc. Similarly, application layer not only provides services to devices and users, but also provides services like data analytics and data mining, etc. Therefore, to provide a flexible and general multi-layer IoT architecture, a service layer is introduced between application layer and network layer for providing the data services. Based on this idea, four-layer SoA has been developed to support IoT [14, 15]. Figure 3 depicts four-layer IoT architecture. The four layers are enlisted as 1. 2. 3. 4.

Perception Layer Network Layer Service Layer Application Layer.

All other three layers of this architecture are similar to layers in three-layer architecture. Thus, service layer is discussed below. Service layer, situated between network and application layer, is responsible for creating and managing the services. In service layer, various enabling technologies are exploited to ensure the efficient service provision. These technologies are interface, resource management and sharing, middleware and service management. This layer relies on the middleware technology for providing functionalities to integrate services and applications in IoT.

Fig. 3 Four-layer architecture

24

Rekha and R. Garg

2.3 Five-Layer Architecture The three-layer architecture describes the basic concept of the Internet of Things. For research, a look into finer aspects of IoT is required, it is not enough for research on IoT. This leads to have many more layered architectures. Another IoT architecture is the five-layer architecture. Figure 4 represents the five layer architecture which consists of perception, transport, middleware, application, and business layers [3, 12]. The five layers are enlisted below. 1. 2. 3. 4. 5.

Perception Layer Transport Layer Processing Layer Business Layer Application Layer.

Perception and application layers have the same role as in three-layer architecture. Hence, role of other three layers is discussed here. 1. Transport Layer: This layer transmits the collected data from lower layer to upper layer securely and vice versa via 3G, 4G, LAN, Bluetooth, NFC, RFID, etc. It also guarantees confidentiality of sensitive information. Thus, this layer is basically responsible for transmitting the information acquired from lower layer to upper layer and vice versa. 2. Middleware Layer: Middleware layer is also known as “Processing Layer”. This layer performs service management and data storage two main functions. Huge amount of information from the lower layer is stored in the database by this layer

Fig. 4 Five-layer architecture

Existing Enabling Technologies …

25

for further data retrieval, processing, and computation. This layer is capable of making decision based on the results of computational automatically. Technologies like databases, big data processing, and cloud computing are employed by this layer. 3. Business Layer: This layer is responsible for managing the whole IoT system, covering both services and applications management. Based on the information received from the lower layer, it can also make business models, reports, flowcharts, graphs, etc. Depending on analysis, it also helps in making more precise decisions about the roadmaps and strategies of business. In addition to this, this layer monitors and manages the underlying four layers.

3 Components of IoT For a better insight of meaning and functionality of IoT, we need to understand the various building blocks of IoT. In this section, we discuss components of IoT required for its functionality. Main six build blocks of IoT are as illustrated in Fig. 5.

3.1 Identification Things, already connected or going to be connected, need to be uniquely addressable. To uniquely identify IoT devices is crucial for IoT success. Various techniques are available for IoT identification like ubiquitous codes (uCode) and electronic product

Fig. 5 IoT components [16]

26

Rekha and R. Garg

codes (EPC) [17]. Furthermore, it is required to differentiate between the device address and its ID for addressing the IoT devices. Device’s ID is name assigned to it. For example, “Wind1” is name for a particular wind sensor. On the other hand, device address represents its address to identify it within a communications network. IPv4 and IPv6 are protocols used for addressing purpose of IoT objects. For making IPv6 addressing suitable for low power consuming wireless networks, 6LoWPAN can be used for compression over IPv6 headers. Addressing is employed to identify devices uniquely because identification schemes are not universally unique. Moreover, devices can use public or private IPs within a network. RFID technology is one of identification methods. RFIDs, coupled with devices act as electronic barcode, are used for the automatic identification of anything.

3.2 Sensing Sensing in IoT system means acquiring data from the environment or other related objects within the network. Gathered data is transmitted to databases, data warehouses on the cloud and analysed for taking the decision required to provide some services. The IoT sensors include smart sensors, actuators, RFID tags or wearable sensing devices. For example, many applications have been developed which enable users to supervise and control a number of smart devices and appliances inside their home, offices, etc., using their smartphones. Sensing devices can be connected to a central management portal to provide the required data by customers.

3.3 Communication Communication technologies in IoT are used to make the connection between the heterogeneous devices so that they can deliver the specific service together. IoT nodes are the constraint nodes, so they should be able to operate consuming low power in lossy and noisy communication links. Various IoT communication protocols are IEEE 802.15.4, Near Field Communication (NFC), LTE-Advanced, Bluetooth, Zigbee, Wi-Fi, Z-wave, RFID, Ultrawide bandwidth (UMB), etc. The communication range, speed, and bandwidth depend on the communication technique used in system. The energy consumption of network also varies with the protocol used for communication. Based on the requirement of the system, suitable protocol can be used for data transfer.

Existing Enabling Technologies …

27

3.4 Computation The data collected by the IoT devices need to be processed for the analysis and decision-making. For this purpose, some processing units (e.g., microcontrollers, microprocessors, SOCs and FPGAs) and software applications are needed which act like the “brain” of IoT and have computational ability. UDOO, Arduino, Mulle Intel Galileo, WiSense, Gadgeteer, Z1, FriendlyARM, BeagleBone, T-Mote Sky, Raspberry PI, Cubieboard, etc. are the several hardware platforms developed to execute the IoT applications. In addition to this, various software platforms are there to support the functionalities of IoT. Operating systems are essential among all software platforms as they enable the execution of all other required software. For example, Riot OS, Contiki, LiteOS and TinyOS are the lightweight OS developed for IoT environments. Cloud platform are also part of IoT computation as they provide facility for big data processing on cloud. There exist some free and commercialized cloud platforms and frameworks to provide IoT services.

3.5 Services IoT provides various types of services. Broadly these services can be classified into four different classes: (i) Identity-related, (ii) Information Aggregation, (iii) Collaborative-Aware and (vi) Ubiquitous Services. Identity-related services are the most essential and basic services because every application has to identify the connected objects. Information Aggregation Services acquire and sum up the raw data which is further processed. On the top of Information Aggregation Services, Collaborative-Aware Services utilize the obtained data for making smart decision and act accordingly. Collaborative-Aware Services can be needed by anyone at anytime. Ubiquitous Services provide these services to anyone anywhere. With this classification, smart grids and smart healthcare belong to the information aggregation class and smart buildings, industrial automation, intelligent transportation systems (ITS), and smart home applications fall into collaborative-aware category.

3.6 Semantics In IoT, semantic means the capability of extracting knowledge smartly via various machines so that the required service can be provided to the users. Knowledge extraction is to discover and utilize the resources and model the information. Further, it also covers recognition and analysis of data for taking smart decision to deliver the

28

Rekha and R. Garg

required service. Hence, semantics refers the brain of IoT by addressing the accurate resource. Semantic Web technologies, like Web Ontology Language (OWL) and Resource Description Framework (RDF) support this requirement.

4 Applications IoT has the potential to make development of a large number applications practically possible. Intelligent applications have been applied in diverse areas as shown in Fig. 6 [18]. But out of these huge number of applications, all are not available to our society yet. However, these applications are capable to enhance the quality of our society. Some examples of IoT applications are smart homes, health monitoring, weather forecasting, smart city, etc. A brief description of diverse IoT application domains is given below.

4.1 Home Automation People are attracted towards the smart homes because of automation. Sensor and actuators with WSN are the reason behind the popularity of home automation. Various

Fig. 6 Different application areas of IoT

Existing Enabling Technologies …

29

Fig. 7 Smart home block diagram

sensors are there in the smart homes to offer the automated and smart services to the end user. These sensors help to automate routine tasks and also in saving the energy by using the motion sensors to switch off the electronic appliances. Moreover, motion sensors can also be exploited for security purpose. Intelligent home applications are genuinely advantageous for the old and differently abled people. Health monitoring system will help to monitor their health and in emergency family/relatives can be informed directly. Further, pressure sensors can be embedded into floors to track the individual’s actions in the smart home. In smart homes, interested events can be recorded by CCTV cameras and features can be extracted to analyse what is going on. Figure 7 illustrates the smart home. Smart home applications also face some challenges and issues. All the events happening in the home are recorded, so security and privacy become important challenges for smart homes. If the security and trustworthiness of the system are compromised by intruder, then the whole system can be made to behave maliciously. As there is no administrator to supervise the system, so reliability is another issue.

4.2 Health care Health care is also taking advantages from the IoT technologies. IoT is providing the benefits to people by monitoring their health regularly, managing their medicines, and also contact with doctors and family in emergency. There are several wearable devices

30

Rekha and R. Garg

exist and are being developed to monitor health of individual’s health condition. Health monitoring applications make possible independent living of oldie. Health applications make independent living possible for oldies and patients having serious health problems. Currently, IoT sensors are utilized to regularly observe and track their health status. In case some abnormality is indicated, warnings can be transferred or can be advised to the patient by IoT application itself depending on the abnormality detected. IoT applications also assist the health system in maintaining an Electronic Health Record (EHR) to track all the medical details of an individual. Healthcare IoT applications can be divided broadly into four groups: tracking, identification and authentication, data gathering, and sensing [2]. There are some IoT applications which help us to stay fit by monitoring our daily routine. For example, fitness trackers are available as wearable devices to measure how many steps are taken. The amount of exercise done by individuals can also be monitored by these trackers.

4.3 Transportation Smart transportation, which helps in managing daily traffic, is implemented by embedding sensors into vehicles, traffic light, roads, etc. These sensors assist in monitoring the vehicles breaching the traffic rules, environment pollution control, traffic congestion reduction, law enforcement, intelligent parking, etc. Block diagram of smart parking system is illustrated by Fig. 8. Smart application help in avoiding the road accidents by spotting drivers who are drunk and also route the traffic properly. Sensors are deployed to conclude the traffic conditions by calculating the speed and distance of the vehicles and pedestrians. This is used to inform the other vehicles. For example, sensors can change the traffic lights when they observe the coming ambulance and let it pass first. These sensors can also inform other lights also. Technologies like GPS, accelerometers, gyroscopes, RFID, cameras, infrared sensors, etc. are used in these applications. There exist various types of applications in transportation.

4.4 Logistics IoT applications also assist in logistics. Real-time information processing system can realize the monitoring of all most supply chain including commodity design, purchasing of raw materials, production, transportation, distribution, sale, etc. NFC and RFID based technologies are used for real-time information processing. Productrelated information can be obtained accurately and timely to assist in responding changeable and complex market in short duration [19]. Railways are the core of any logistics. Smart identification system based IoT has been introduced for effective logistics management [20].

Existing Enabling Technologies …

31

Fig. 8 Smart parking system

4.5 Smart Environment and Agriculture Agricultural production is affected by environmental parameters like humidity, soil information and temperature. A group of sensors can be deployed by farmers to identify various land requirements. This information can be used to take the necessary action for efficient production. IoT-based agriculture enables us to study the real production situation while managing from remote area. Smart farming assists in adopting effective farming actions by getting conversant with different possible climate and land conditions. Smart wrapping up of seeds and fertilizers to provide according to specific environmental conditions are some use cases of smart farming. This will extremely ameliorate the agricultural productivity by avoiding the improper farming practices. Today, air pollution is a main concern due to climate change of the earth and degradation of air quality. Vehicles are responsible for a lot of air pollution. IoT based applications help in monitoring and controlling the air pollution. To measure the air pollution, electrochemical toxic gas sensors can be used. RFID readers can be deployed on both sides of the road to identify the vehicles along with the gas sensors. This will help in taking action against vehicle polluting the air.

32

Rekha and R. Garg

IoT applications are not only limited to the domains discussed above. There are many more areas in which IoT applications are being deployed everyday. Other application sectors are Natural Calamities Prediction [21], Public Safety, Smart Watering System, Infrastructure Monitoring, Aquaculture, Smart Society, Smart City, Urban Management, Smart Cycling, Smart Sports, Smart Tourism, Cloud Service and Management, Augmented Maps, Industrial Plants, Smart Grid, Smart Meter, Building Management, Defence, etc. [22].

5 Challenges in IoT Although there are many enabling technologies which make IoT conception practical, still it requires a lot of research efforts. IoT encounters several challenges which need to be attended for better performance. Some IoT issues are as follows: 1. Security: The IoT is highly susceptible to various attacks due to several reasons like wireless communication, unattended components, low capabilities in terms of both computing resources and energy. According to Balte et al.[23], IoT faces some security challenges which are Privacy, Authentication, Trust, Confidentiality, Policy enforcement, Access control, etc. Security challenges encountered at different layers of IoT architecture are [15] a. Perception Layer: As the perception layer focuses on data collection, security challenges at this layer are Node Capture, Eavesdropping and Interference, False Data Injection, Cryptanalysis and Side Channel, Malicious Code Injection, Replay Attacks, Sleep Deprivation Attacks, etc. b. Network Layer: As the network layer focuses on data transmission, security challenges at this layer are related to availability of network resources and wireless networks. These are Denial-of-Service (DoS) Attacks, Man in the Middle Attack, Wormhole Attacks, Sinkhole Attacks, Spoofing Attacks, Routing Information Attacks, Sybil Attacks, Unauthorized Access, etc. c. Application Layer: As the application layer focuses on service support, challenges faced are related to software attacks. These are Phishing Attack, Malicious Scripts and Virus/worm, etc. 2. Standardization: Scientific community has made many contributions for the standardization and complete deployment of IoT framework. There are various standards for IoT like Zigbee, Wireless Hart, M2M, NFC, ROLL, etc. But these are not integrated in an extensive framework [24]. 3. Addressing: Number of IoT devices are increasing rapidly. To identify these large number of devices, effective addressing policies are required. Currently, the IPv4 protocol is used for addressing purpose through a 32-bit address. Available IPv4 addresses are exhausting rapidly. Hence, other addressing schemes should be adopted. Further, way of obtaining address is also an issue. Object Name Servers

Existing Enabling Technologies …

33

(ONS) are required to associate a reference to a description of a specific object and the allied identifier, and vice versa. 4. Transport Protocols: A new concept of the transport layer is also needed for the IoT. Transport layer aims at providing guaranteed end-to-end reliability and to control end-to-end congestion. Due to congestion control and connection setup, existing transport protocols do not perform well in the IoT scenarios. Hence, TCP protocol cannot be used effectively in the IoT. 5. Traffic characterization and QoS support: A huge amount of traffic will be generated by IoT having various patterns. These patterns are expected to be notably divers from which are discovered in the present Internet. Therefore, traffic characterization and modelling along with traffic requirements are required so that accurate solutions can be devised to support quality of service. 6. Energy Efficiency: The key characteristic of IoT devices is the power limitation. Many sensors are powered via battery without recharging. Frequent battery replacement for the sensors exploited in the outdoor territories can cause to a significant expenditure. IoT devices are growing rapidly and this rapid increment of IoT is causing significant energy consumption [25]. Therefore, Energy efficiency (EE) is a critical issue for the large scale deployment of IoT and WSN applications [26]. Some major IoT challenges are discussed above. In addition to the these major challenges, IoT also faces some other challenges. These are GIS-based visualization, data analytics, cloud computing, participatory sensing, data mining and management, data communication protocols, networking and storage, forensics challenges, etc. [27–29]. For proper functioning IoT device requires power. These devices are mostly powered by batteries. But, these devices are size constrained which also put limits on battery size. Battery has the limited energy capacity which restricts the device’s operations and lifetime. So, it is needed to replace the battery so that the device can operate. Many applications need the deployment of devices in such an area where battery replacement is very difficult. Hence, energy provision is a critical issue for the IoT devices.

6 Energy Management In the future, sensor deployment will be enormous. A sensor node consists of: (i) Sensing subsystem (ii) Processing subsystem (iii) Communication subsystem (iv) Power source. Among all these subsystems, the communication subsystem requires more energy on average than the other subsystems [30]. Expansion of IoT offers a lot of opportunities, but it also has some risks. Servicing cost will be a major issue for these sensors. Therefore, economical deployment and maintenance of these sensors is one challenge. Most of IoT devices are battery driven. Moreover, battery replacement is also important issue for sensors. It is not impossible to replace the

34

Rekha and R. Garg

battery after deployment, but it is practical impossible. Hence, low power consuming sensor designs or techniques are required for energy efficiency. For example, sensor’s lifetime should be more than the animal’s when sensor is exploited in animal for tracking purposes. Even IoT helps in reduction of power consumption [31, 32], but we need to reduce the power consumption of IoT devices. According to Cisco, in 2008/2009, the number of devices connected to Internet surpassed the number of humans [33]. In future, this number will increase. Many applications require sensor to be positioned into unwelcome territories. For such type of application, battery replacement is a tedious job and even sometimes it costs more than the device’s cost. Despite these restrictions, applications need power for functioning, but batteries have short lifetime. So, providing power to these growing number of objects is a major challenge. Energy is one of the scantiest resources in sensor networks. So it is one of the most critical issue in IoT. In applications like data gathering, environmental monitoring, tracking, etc., using a fixed power source and/or recharging a battery manually may not be both economically and technically feasible. To solve this problem, we need to manage energy in IoT devices. Energy management can help in prolonging the lifetime of device. Energy management can be defined as combination of two set of principles to manage several energy supplying mechanisms and further effective consumption of the supplied energy in a device. It is required for a node to have an effective energy management technique for the limited power source along with meeting the application requirement considering the available energy source. Thus, energy management can be done via energy harvesting or energy conservation. Both of these techniques are promising solution for the IoT power issue. Both of these techniques are discussed below. 1. Energy Harvesting: In this technique, ambient energy is scavenged and converted into useable form. There are many types of energy available in our surrounding like solar, wind, hydro, etc. These energy sources are replenishable. 2. Energy Conservation: Another technique is energy conservation which focus on the reduction of energy consumption in IoT devices. Node lifetime can be enhanced, if the energy usage can be deducted. There are many schemes to conserve the energy like duty cycling, data reduction, etc. This section focuses the energy management in IoT. Various techniques to manage the energy in IoT devices are discussed here.

6.1 Energy Harvesting Energy harvesting, which is also known as energy scavenging or power harvesting, is the process used to derive the energy from external sources and can be stored for small, wireless autonomous devices used in wireless sensor networks. Energy harvesting techniques are used to acquire energy from the renewable resources. Energy harvesting device transfer the energy into electrical energy for the use. It can also be

Existing Enabling Technologies …

35

Fig. 9 Block diagram of energy harvesting system

stored in the storage devices for the later use. Figure 9 illustrates the simple block diagram of energy harvesting system. The component of an energy harvesting system is [34]: 1. Energy Source: For harvesting the energy, an ambient energy source is required from which energy is collected. There exist various ambient energy sources which are renewable. 2. Energy Collecting Module: This module obtains the energy from the surrounding and converts it into useable form. a. Sensing Module: Ambient energy needs to sensed for harvesting. Depending on the type of energy to be harvested, energy sensing module is required. For example, solar panel is required to collect the solar energy. b. Controlling Circuit: After harvesting the energy, it needs to be converted into electricity before use. Oscillator and rectifier circuits can be used for converting the energy into AC and DC respectively depending on application. After this, electrical power need to be regulated by the internal circuitry. 3. Storage Device: This power can be delivered to load directly or via storage device. If harvested energy needs to be used later, then it needs to be stored. All the time, it is not possible to use the harvested power at the same time. For example, solar energy is not available in the nights. So it needs to be stored for night usage. There are many storage devices which can be used depending on the application requirements. Flywheels, Hydrogen, Compressed-Air Energy Storage (CAES), Super-capacitors, Superconducting Magnetic Energy Storage (SMES), Pumped Hydroelectric Energy Storage (PHES), Battery Energy Storage (BES), etc., are some examples of storage devices. 4. Load. Ambient energy is the base of the energy harvesting concept. Naturally, replenished energy sources are called renewable energy sources. There exist various ambient energies which can be harvested to power the devices. But the amount of converted electrical energy is not same as amount of available energy. This depends on the sensing module and converting circuit. The energy conversion efficiency varies from one circuit to another depending on the technology used. Various renewable energies are shown in Fig. 10.

36

Rekha and R. Garg

Fig. 10 Various ambient energies

Solar Energy: Sun is source of huge amount of solar energy. Solar energy is extensively available during the whole sunny day. There are various technologies used to collect the solar energy. These technologies are photo-voltaic (PV) panels, concentrating photo-voltaics (CPV) and concentrating solar thermal power (CSP). PV panels are the most used technique among all three. CPV and CSP are macro power generating system while PV is micro power generating system. So, PV is used in sensor network applications. Solar energy is not available during cloudy days and nights. Therefore, solar power needs to be stored during cloudy days and nights whenever required. Solar to electrical power conversion efficiency is low. Additionally, hardware part of solar harvesting system causes energy loss. Therefore, it results in low efficacy. Solar energy is most appropriate in terms of power density and power wireless

Existing Enabling Technologies …

37

sensors. Solar energy is mostly suitable in outdoor environment. But it may be beneficial for some indoor environment which have sufficient sunlight like stadiums, hospitals, etc. [30]. Wind Energy: Wind energy is worldwide available and fastest growing renewable energy. It is one of the most economical suitable techniques. It also generates power without causing any pollution. Hence, wind energy has a great potential as a comprehensive clean renewable energy resource. Wind turbines are used to generate the electricity as the speed of the wind cause the motion in blades of turbines. Various wind turbine technologies exist to obtain the power from wind energy. These technologies are Variable-Speed Concept Utilizing Full-Power Converter, Variable-Speed Concept Utilizing Doubly Fed Induction Generator (DFIG) and Semiconductor-Device Technology [35]. Vibrational Energy: It is also known as mechanical energy. Sources of vibrational energy are vibrations, stress–strain, pressure, etc. Appropriate Mechanical-toElectrical Energy Generator (MEEG) is required to convert the mechanical energy into electrical. Various mechanisms used by MEEG are electrostatic, electromagnetic and piezoelectric mechanisms. Electrostatic or piezoelectric generators are used to convert the pressure variation into power which offer the highest power density. These techniques are adapted extensively. Hydropower Energy: Hydropower energy is the biggest and one of the oldest renewable source used to produce electricity. Electricity is generated by controlling the force of water flow in dams. Conversion efficiency of modern hydro turbines is upto 90%. At present, many small commercialized (350–1200 W) units can be deployed in rivers and streams. Energy cannot be harvested only from water but also be harvested from moving liquids using a small hydro-generator. Thermal Energy: Thermal energy harvesting uses the Seebeck effect. Temperature difference between two surfaces is used to produce the electricity. Heat is generated as a by-product of several processes. Many applications execute process working on the principle of temperature difference. Hence, thermal energy can be considered as rich source of energy. There exist many large and small scale devices to generate electrical power from thermal energy. Biomass Energy: Biomass includes all organic material produced by trees, plants, crops, animals, etc. Crops, wood, manure, fruits, etc. are some examples of biomass fuels. It accumulates the solar energy through photosynthesis. Bioenergy can be converted into electricity, heat, liquid fuels and gases like methane. Geothermal Energy: Earth’s core is source of abundant energy which is resulted from the ancient heat, sliding of continental plates and decomposition of radioactive elements. Electricity and heat can be generated by digging deep wells and withdrawing the underground heated water to the surface or geothermal plants. Geothermal energy is available in tremendous amount. It has been generated commercially for last 70 years for direct use or electricity generation. Human-based Energy: Sensors are extensively deployed in health care systems. Sometimes, these are even exploited into human body. Due to sensor deployment in human body, they need to be active for longer duration. Therefore, energy harvesting from the human itself is given preference than the other power sources. Human-based

38

Rekha and R. Garg

energy can be collected in a different ways like through body heat, heartbeat, locomotion, blood flow or change in finger position. AM Signal Energy: For almost last 100 year, AM signals have been exploited to power the crystal radios without any external batteries [34]. It has been proved by the researchers that AM signals can be considered as a more effective renewable energy source. Three main ambient sources solar, thermal and vibrational energy may or may not provide energy continuously throughout the whole day. But this problem can be solved by the AM signals that are broadcast 24 hours. These signals will never reach to zero but decay with time. Soil Energy: Sensors are also deployed in the agriculture area to enable the smart farming. These sensors can be powered by exploiting the soil energy. There exist microorganisms, also known as exoelectrogens, in the soil which can collect the soil energy. Conversion efficiency of soil cell varies with varying operative conditions. Renewable energy sources are promising solution to solve the powering problem of IoT devices. Availability and feasibility should be considered while designing energy harvesting algorithms. Researches have developed many energy harvesting techniques. They are also focusing on increasing the conversion efficiency of different energy convertor. Thus, energy harvesting achievements appear to be promising toward the deduction of energy insufficiency in the sensor networks.

6.2 Energy Conservation In almost cases, it is possible to acquire the power from the environment by using solar energy, thermal energy, etc. Although these external sources often do not provide continuous behaviour. Thus, energy buffering is necessary. Hitherto, energy is very scare resource and it must be utilized very sparingly in any case. Therefore, it is required that available energy must be used effectively. Hence, energy conservation is one of key issue in sensor networks. Many approaches have been developed by researchers to conserve the energy by the nodes. These schemes can be classified broadly into the following three categories: 1. Sleep/Wake-up 2. Data Driven 3. Mobility Based. Figure 11 represents the high level taxonomy of various energy conservation schemes.

6.2.1

Sleep/Wake-up Scheme

Sensor node does not receive or send data every time. There are some state slots when these nodes do not communicate. These are called idle states. Nodes also consume

Existing Enabling Technologies …

39

energy in these states. Thus, idles states are the main reasons of energy consumption in the nodes. Idle listening of wireless transceiver requires equal amount of energy when it sends or receives and it is higher than the consumption of sleep state. Energy consumption can be reduced by putting to node at low energy consumption mode. During the idle states, nodes can go to sleep state so that they require less power. It reduces the unnecessary consumption of energy. Switching between the sleep and wake-up states is the main concept of sleep/wake-up schemes. The node is switched On when it communicates and goes to Off during inactive period. These techniques adjust the radio state of the node to cut down the energy requirement of the node. These techniques can be categorized as follows: 1. Duty Cycling: It is the fundamental approach in most of the wireless sensor network to enhance the lifetime of node. These schemes schedule node’s state based on the network activity to understate idle listening. Duty cycle of a node is fraction of time for which node remains active during its lifetime. These schemes can be divided into three categories: synchronous, asynchronous and semi-synchronous [37]. Duty cycling schemes are surely quite energy-efficient, however these approaches have high sleep latency as a node has to wait for the receiver to be in active state. Moreover, in some cases, a node cannot broadcast information to all of its neighbouring node because they are not awake simultaneously. 2. Passive wake-up radios: In the duty cycling, node switches between inactive and active states according to their duty cycle. There can be unnecessary wake-up calls even though there is no communication. So these undesirable wake-ups waste the energy in duty cycling. To resolve this issue, passive wake-up radio are used to reduce the energy consumption. Passive wake-up radios are low power consuming radio which awake the node when data transmission is required. In these schemes, two types of radios are used. First one is low power consuming passive radio which active the node and second one is energy hungry radio used for transmission. Topology control can be done based on the connectivity or location. 3. Topology control: To find out the optimal subset of nodes which guarantee connectivity is known as the topology control. In this technique, network redundancy is exploited to increase the network’s lifetime. Sensors are deployed randomly. It may be possible to deploy more sensors than the required. Then it is possible to deactivate some node at a time and active them later to prolong the life of network. Hence, topology control focuses on adapting network topology dynamically according to the application requirement. 4. MAC Protocol: Medium Access Control (MAC) protocols address two important issues: resource sharing and multiple access control. Collision and control packet overheads are also reasons for the energy deficiency in the sensor networks. Thus, several MAC protocols have been proposed with low energy consumption. MAC protocols can be classified into three categories given as: Contention based, TDMA based and hybrid protocols.

Fig. 11 Taxonomy of energy conservation schemes [36]

40 Rekha and R. Garg

Existing Enabling Technologies …

41

These techniques have potential to save the energy of node. Besides their efficiency, these impact on latency of data gathering as node must be woken up to receive the data. It is also difficult for a node to connect to its neighbours before communication at the same time because they do not have same wake-up schedule. Many efforts have been done to enhance the performance of these techniques in terms of latency and memory overflow.

6.2.2

Data-Driven Scheme

Above-discussed sleep/wake-up techniques are typically unaware of data to be sampled by the nodes. Energy consumed by a node increases with the amount of data to transmitted and received by the node. Therefore, data-driven schemes are used further to ameliorate the energy efficiency. These approaches aim at the data reduction to be communicated for saving the energy. Energy consumption of a node is affected by the data sensing in the following two ways: • Power consumption of sensing unit may not be negligible. • Redundant sampling requires the energy while sampling the data having strong temporal and/or spatial correlation. So, data driven approaches are used to limit sensing tasks and the unneeded sampling because both the data acquisition and transmission are expensive in terms of energy. These approaches specifically focus on reducing the amount of data to be sampled while considering sensing accuracy. Such approaches can be divided into two classes: data reduction and data acquisition schemes. 1. Data Reduction: These approaches address the issue of redundant data samples. Such schemes aim at reducing the data to be communicated to sink node. a. Data Compression: Data compression is the process of modifying, converting or encoding the data in such a way that the new representation requires less space than the original. Data compression is also well-known as bit-rate reduction or source coding. Hence, these approaches reduces the data to be transmitted by compressing it and help in energy conservation. These are very energy efficient as they shorten the packet size. However, already existing compression techniques are not suitable to resource constraint sensor nodes. Hence, it is required to develop compression approaches particular to sensor nodes. b. Data Prediction: These techniques are based on the building a prediction model upon the sensed phenomenon. Obtained model resides both at the sink node and source node. It is used to predict the sensed value by the node within certain error range. If user query is satisfied with a certain level of accuracy, then sink node evaluates the query without getting the exact

42

Rekha and R. Garg

data from the source node. But if it is not satisfied, explicit communication takes place between the source and sink node. Thus, these techniques assist in reduction of energy consumption by decreasing the number of communication between sink and source node. Data prediction approaches can be divided further into stochastic, time series forecasting and algorithmic approaches. c. Data Aggregation: In these techniques, data fusion take place along the path from sources to sink node to reduce the amount of data to be transmitted. An intermediate node between the sources and sink node performs the data aggregation. For example, a node receives data from several sources. But it forwards only aggregated value (like maximum, minimum or average value of all the received data). It serves the purpose of data reduction along with improved latency but accuracy may be compromised. 2. Data Acquisition: Only reducing the data communication is not sufficient, energy saving schemes also need to focus on the data sampling rate. This group of classification focuses on the reduction of energy consumption of node’s sensing unit. These protocols consider that significant amount of energy is required by the sensing unit compared to communication system. Such approaches aim at data sampling for data collection. Regularly sampled data can follow a pattern or it can have spatial or temporal or both correlation. Thus, these techniques aim at reducing the number of data acquisition. These schemes can be grouped as follows: a. Adaptive Sampling: This group of techniques exploit the temporal and spatial correlation of data to reduce the data sampling rate. Hence, these approaches help in reducing the energy consumption due to sampling rate. These approaches are more efficient and general. Mostly, these are enforced in a centralized fashion because they require high computations. b. Hierarchical Sampling: This group assumes that the network consisting of various kinds of sensor. Each sensor has different resolution and energy consumption. These schemes select which sensor class to activate dynamically to find out a trade off between energy conservation and accuracy. These are more application specific. c. Model-based Active Sampling: These schemes are based on prediction. A model is built using the sampled data which is used to forecast the future values within a certain level of accuracy. These techniques exploit this model to cut down the frequency of data sampling which assist in reduction of energy consumption. Data-driven approaches are very promising as they are quite general, efficient and feasible. More energy-efficient approach can be a combination of both spatial and temporal relation so that multiple direction of data redundancy can be exploited.

Existing Enabling Technologies …

6.2.3

43

Mobility Based Scheme

Almost all the approaches assume that nodes in the network are static. But recently, mobility of node has been considered for energy-efficient data gathering in sensor networks. These approaches take mobility on node into consideration. Mobility of node is feasible and it can be achieved in various ways. For example, mobilizer can be equipped with the sensor to change its location, but mobilizers are quite costly from the energy consumption point of view. So adding mobilizer is not a good idea. But limiting the mobility to a limited number of energy constraint nodes can be beneficial. On the other hand, instead of using mobilizer, sensors can be deployed into mobile elements like car, animals, bus, etc. In this case, there is no extra energy consumption but mobility pattern of mobile element need to be considered. Mobile nodes can be part of network or it can be part of environment. When a mobile node is part of network, it can be controlled completely. In another case when node is part of environment, it might not be possible to control it. Node mobility is quite useful for decreasing energy consumption. Packets from source to sink traverse network following multi-hop path. When the nodes are static, some nodes on a particular path are more loaded than the others according to network topology used. Specially nodes near to sink node consume more energy because they have to receive and forward each packet directed towards sink node. Therefore, these nodes will have energy deficiency. Thus, mobile nodes are necessary to prolong network lifetime. Mobility-based approaches can be categorized based on whether relay node is mobile or sink node is mobile. Hence, these approaches can be grouped as follows [36]: 1. Mobile Relay: These techniques are based on mobility of relay nodes. Mobile relay model is already explored for data gathering in multi-hop networks. In this model, relay nodes move across the network and collect the data from source node. The mobile relay node carry the acquired data and forward it towards the sink node. Mobile relay nodes are also known as message ferries. Hence, message ferries can be observed as a moving communication infrastructure which facilitate data transmission in distributed wireless networks. 2. Mobile Sink: In these techniques, a sink is considered to be mobile which acquire the data from source nodes in the network so that network lifetime can be increased. In mobile sink model, mobile sink moves to a limited number of positions. Mobile sink node stays at these locations for a time duration and collects the data the nodes. If some nodes are not in coverage area of mobile sink node, then they can transfer data using multi-hop path. It has been observed that by using mobile sink nodes network’s lifetime can be improved by 5–10 times than static sink nodes. However, it can affect the latency related to data received at the sink node. Recently, research related to mobile data collector has provided possible solutions to obtain energy-efficient data gathering approaches. Various methods have been introduced for the movement of mobile relay and mobile sink node. These techniques

44

Rekha and R. Garg

help in prolonging the network lifetime by saving energy of nodes near to sink nodes. However, many aspects need to be considered like timely discovery of mobile elements by static nodes, efficient data transmission between mobile mode and static node, etc.

7 Research Directions Various open challenges in IoT have been discussed already. These challenges include security, privacy, authentication, transmission protocols, standardization, addressing, quality of service, energy efficiency, etc. These all challenges provide room for further improvement in IoT. Power provision to nodes is still one of the most critical issue. Many techniques have been developed for sensor to make them energy-efficient. But these techniques are still open for further enhancement. Energy harvesting from ambient environmental sources is still under improvement and hence poses many challenges. It is still an open research issue. It requires a lot of efforts towards the realization of a perpetual IoT. In addition, energy harvesting efficiencies for already existing energy resources can be improved. Researchers also advocate exploring new sources of energy for small powered nodes. A hybrid approach comprising the existing sources (batteries and ambient environment) can also be enforced in order to enhance the network lifetime. Most of the energy harvesting techniques harvest the energy from single energy source which may or may not sufficient for the node’s operation. So, it is required to think about energy harvesting from multiple sources. For the network-wide energy management, it requires knowledge of both energy supply and consumption for maintaining an energy effective network operation. Hence, it is required to consider both power supply as well its consumption parallelly to design an energy-efficient algorithm. Energy conservation schemes are promising solution for the energy deficiency in sensor node. These schemes assist in increasing the network lifetime. These approaches conserve the energy but latency and data accuracy can be affected. Therefore, energy conservation schemes with the improved latency and data accuracy are needed. Energy consumption can be further reduced by combining more than one conservation techniques. Hybrid techniques can assist in energy management. For example, data compression can applied with duty cycling approach.

8 Conclusion This chapter provides an introduction to “Internet of Things” (IoT) and its energy efficiency. IoT is emerging technique which is making the world connected. Various researchers defined IoT according to their point of view. Some of IoT definitions have been discussed here. There are various architectures for IoT as there is no single universally agreed architecture. IoT has six main components which are necessary to

Existing Enabling Technologies …

45

understand for gaining the meaning of IoT functionality. These components include identification, sensing, communication, computation, services and semantics. IoT has been exploited in various domains. It has a wide area of application which include healthcare, safety, smart environment, etc. Some of IoT applications have been explained in this chapter. Besides the popularity of Iot, it also has many issues like security, addressing, visualization, energy management, etc. Energy deficiency is a critical issue for the IoT device which needs to be resolved. Many energy management techniques have been discussed which address the issue of energy deficiency in IoT. Energy management techniques can be classified into two classes given as: energy harvesting and energy conservation. This chapter provides the taxonomy of various energy harvesting and energy conservation approaches with the future directions.

References 1. Santucci, G., et al.: From internet of data to Internet of Things. In: International Conference on Future Trends of the Internet, vol. 28 (2009) 2. Atzori, L., Iera, A., Morabito, G.: The Internet of Things: a survey. Comput. Netw. 54(15):2787– 2805 (2010). https://doi.org/10.1016/j.comnet.2010.05.010, http://www.sciencedirect.com/ science/article/pii/S1389128610001568 3. Tuwanut, P., Kraijak, S.: A survey on IoT architectures, protocols, applications, security, privacy, real-world implementation and future trends. In: 11th International Conference on Wireless Communications, Networking and Mobile Computing (WiCOM 2015), p. 6 (2015). https://doi.org/10.1049/cp.2015.0714, http://digital-library.theiet.org/content/conferences/10. 1049/cp.2015.0714 4. ITU: Internet of Things Global Standards Initiative. https://www.itu.int/en/ITU-T/gsi/iot/ Pages/default.aspx. Accessed 17 Aug 2018 5. Srivastava, L.: Pervasive, ambient, ubiquitous: the magic of radio. In: European Commission Conference From RFID to the Internet of Things, Bruxelles, Belgium (2006) 6. Van Kranenburg, R.: The Internet of Things: a critique of ambient technology and the all-seeing network of RFID. Institute of Network Cultures (2008) 7. Infso D: Internet of Things in 2020: a roadmap for the future, INFSO D. 4 networked enterprise & RFID and INFSO G. 2 micro & nanosystems in co-operation with RFID working group of the European technology platform on smart systems integration (eposs). European Commission, Brussels, Belgium, Technical Report (ver 3) (2008) 8. Vermesan, O., Friess, P., Guillemin, P., Gusmeroli, S., Sundmaeker, H., Bassi, A., Jubert, I.S., Mazura, M., Harrison, M., Eisenhauer, M., et al.: Internet of Things strategic research roadmap. Internet Things Glob. Technol. Soc. Trends 1(2011), 9–52 (2011) 9. Peña-López, I., et al.: ITU Internet report 2005: the Internet of Things (2005) 10. Gartner: http://www.gartner.com/newsroom/id/2905717. Accessed 17 Aug 2018 11. Scanner V: https://www.venturescanner.com/internet-of-things. Accessed 17 Aug 2018 12. Sethi, P., Sarangi, S.R.: Internet of Things: architectures, protocols, and applications. J. Electr. Comput. Eng. (2017) 13. Zhu, Q., Wang, R., Chen, Q., Liu, Y., Qin, W.: IOT gateway: bridging wireless sensor networks into Internet of Things. In: Proceedings - IEEE/IFIP International Conference on Embedded and Ubiquitous Computing, EUC 2010, pp. 347–352 (2010). https://doi.org/10.1109/EUC. 2010.58 14. Da Xu, L., He, W., Li, S.: Internet of Things in industries: a survey. IEEE Trans. Ind. Inform. 10(4), 2233–2243 (2014)

46

Rekha and R. Garg

15. Lin, J., Yu, W., Zhang, N., Yang, X., Zhang, H., Zhao, W.: A survey on Internet of Things: architecture, enabling technologies, security and privacy, and applications. IEEE Internet Things J. 4(5), 1125–1142 (2017) 16. Al-Fuqaha, A., Guizani, M., Mohammadi, M., Aledhari, M., Ayyash, M.: Internet of Things: a survey on enabling technologies, protocols, and applications. IEEE Commun. Surv. Tutor. 17(4), 2347–2376 (2015) 17. Koshizuka, N., Sakamura, K.: Ubiquitous ID: standards for ubiquitous computing and the internet of Things. IEEE Pervasive Comput. 4, 98–101 (2010) 18. Ray, P.P.: A survey on Internet of Things architectures. J. King Saud Univ. Comput. Inf. Sci. 30(3), 291–319 (2018). https://doi.org/10.1016/j.jksuci.2016.10.003 19. Yuan, R., Shumin, L., Baogang, Y.: Value chain oriented RFID system framework and enterprise application. Beijing: Science (2007) 20. Guo, Z., Zhang, Z., Li, W.: Establishment of intelligent identification management platform in railway logistics system by means of the Internet of Things. Procedia Eng. 29, 726–730 (2012) 21. Singh, R., Gehlot, A., Samkaria, R., Mittal, M.: Iot based intelligent robot for various disasters monitoring and prevention with visual data manipulations. Int. J. Tomogr. Simul. 32(1), 90–99 (2019) 22. Singh, R., Gehlot, A., Mittal, M., Samkaria, R.: Application of icloud and wireless sensor network in environmental parameter analysis. Int. J. Sens. Wirel. Commun. Control. 7(3), 170–177 (2018) 23. Balte, A., Kashid, A., Patil, B.: Security issues in Internet of Things (IoT): A survey. Int. J. Adv. Res. Comput. Sci. Softw. Eng. 5(4) (2015) 24. Li, S., Da Xu, L., Zhao, S.: 5G Internet of Things: a survey. J. Ind. Inf. Integr. (2018) 25. Ji, B., Song, K., Li, C., Zhu, W.P., Yang, L.: Energy harvest and information transmission design in internet-of-things wireless communication systems. Int. J. Electron Commun. (AEÜ) 87, 124–127 (2018). https://doi.org/10.1016/j.aeue.2018.01.038 26. Nguyen, T.D., Khan, J.Y., Ngo, D.T.: Energy harvested roadside IEEE 802 . 15 . 4 wireless sensor networks for IoT applications. Ad Hoc Netw. 56, 109–121 (2017). https://doi.org/10. 1016/j.adhoc.2016.12.003 27. Bandyopadhyay, D., Sen, J.: Internet of Things: Applications and challenges in technology and standardization. Wirel. Pers. Commun. 58(1), 49–69 (2011). https://doi.org/10.1007/s11277011-0288-5. 1105.1693 28. Gubbi, J., Buyya, R., Marusic, S., Palaniswami, M.: Internet of Things (IoT): a vision, architectural elements, and future directions. Futur. Gener. Comput. Syst. 29(7), 1645–1660 (2013). https://doi.org/10.1016/j.future.2013.01.010. 1207.0203 29. Conti, M., Dehghantanha, A., Franke, K., Watson, S.: Internet of Things security and forensics: challenges and opportunities (2018) 30. Shaikh, F.K., Zeadally, S.: Energy harvesting in wireless sensor networks: a comprehensive review. Renew. Sustain. Energy Rev. 55, 1041–1054 (2016). https://doi.org/10.1016/j.rser. 2015.11.010 31. Reka, S.S., Dragicevic, T.: Future effectual role of energy delivery: a comprehensive review of Internet of Things and smart grid. Renew. Sustain. Energy Rev. 91, 90–108 (2018) 32. Moghimi, M., Rafi, F.H., Jamborsalamati, P., Liu, J., Hossain, M., Lu, J.: Improved unbalance compensation for energy management in multi-microgrid system with Internet of Things platform. In: 2018 IEEE International Conference on Environment and Electrical Engineering and 2018 IEEE Industrial and Commercial Power Systems Europe (EEEIC/I&CPS Europe). IEEE, pp. 1–6 (2018) 33. CISCO: How the Internet of Things will change everything. https://blogs.cisco.com/ digital/how-the-internet-of-things-will-change-everything%E2%80%94including-ourselves. Accessed 17 Nov 2018 34. Ferdous, R.M., Reza, A.W., Siddiqui, M.F.: Renewable energy harvesting for wireless sensors using passive RFID tag technology: a review. Renew. Sustain. Energy Rev. 58, 1114–1128 (2016)

Existing Enabling Technologies …

47

35. Carrasco, J.M., Franquelo, L.G., Bialasiewicz, J.T., Member, S., Galván, E., Guisado, R.C.P., Member, S., Ángeles, M., Prats, M., León, J.I., Moreno-alfonso, N.: Power-electronic systems for the grid integration of renewable energy sources: a survey. Ieee Trans. Ind. Electron. 53(4), 1002–1016 (2006). 1006.5277 36. Anastasi, G., Conti, M., Di Francesco, M., Passarella, A.: Energy conservation in wireless sensor networks: a survey. Ad Hoc Netw. 7(3), 537–568 (2009) 37. Carrano, R.C., Passos, D., Magalhaes, L.C., Albuquerque, C.V.: Survey and taxonomy of duty cycling mechanisms in wireless sensor networks. IEEE Commun. Surv. Tutor. 16(1), 181–194 (2014)

Energy-Efficient System Design for Internet of Things (IoT) Devices Neeta Singh, Sachin Kumar, Binod Kumar Kanaujia, Hyun Chul Choi and Kang Wook Kim

Abstract Nowadays, the Internet of Things (IoT) technology is increasing vastly and in the coming days, billions of things/devices/systems will be connected with each other through the internet. For making this emerging technology self-sustainable, there is a need for green and incessant energy, to power up different nodes of the IoT system, and it is only possible with the use of energy harvesting schemes. A number of renewable resources are available in nature and by using some means and methods the available renewable energy can be harnessed effectively. In this chapter, a brief overview of different energy harvesting schemes is presented along with their advantages and limitations. The wireless power transfer and wireless energy harvesting scheme using rectenna technology are also discussed in detail in the subsequent section. In the last subsection, the latest applications of IoT with their influence on human life are discussed. Keywords Energy harvesting · Green energy · IoT · Wireless transmission · WSN

N. Singh Faculty of Engineering and Technology, Department of Electronics and Communication Engineering, Jamia Millia Islamia, New Delhi 110025, India e-mail: [email protected] S. Kumar · H. C. Choi · K. W. Kim (B) School of Electronics Engineering, Kyungpook National University, Daegu 41566, Republic of Korea e-mail: [email protected] S. Kumar e-mail: [email protected] H. C. Choi e-mail: [email protected] B. K. Kanaujia School of Computational and Integrative Sciences, Jawaharlal Nehru University, New Delhi 110067, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. Mittal et al. (eds.), Energy Conservation for IoT Devices, Studies in Systems, Decision and Control 206, https://doi.org/10.1007/978-981-13-7399-2_3

49

50

N. Singh et al.

Fig. 1 Block diagram of M2M communication

1 Introduction Machine to Machine (M2M) communication refers to the conversation between two electronic devices without the assistance of humans. Figure 1 illustrates the block diagram of M2M communication describing how one machine communicates with other machines with the help of Internet of Things (IoT) [1–4]. A communication network connects the M2M gateway and application devices. This is the link through which Radio Frequency Identification (RFID) and Wireless Sensor Networks (WSN) are connected using the internet or radio signal with the M2M nodes. These links can be wired or wireless depending upon the availability, hence providing a medium through which several M2M devices can operate concurrently. For instance, M2M application in healthcare, in this case, a chip is placed inside the human body or on the body if it is a wearable circuit consisting of electrocardiogram (ECG) sensor, motion sensor, body temperature sensor, blood sugar sensor, etc. If the sugar level of the body increases above the prescribed limit, a notification is received on the person smartphone with the help of the Internet and further transmitted to the hospital or the concerned doctor.

2 Operation IoT connects millions of systems and devices through a common network. The devices connected may be home appliances, human body, vehicles, buildings, streetlight, healthcare mechanisms, etc. M2M communication allows people to access the information about the surroundings with the help of a sensing element which gathers the information and transmits the data to the network; the network may be a guided

Energy-Efficient System Design for Internet …

51

Fig. 2 Schematic of the water level sensor node

medium or unguided media, considering low cost and easy accessibility [5, 6]. The physical quantity in the environment is detected by the help of a sensor which may be an electromechanical device. The sensor captures the desired signal, converts the sensed data in the preferred form, and transmits the information to the subsequent devices/system. Few types of sensors commonly used these days are smartphones with touchpads, touchscreen notebooks, motion sensor in automatic gates, photo sensors to detect infrared, ultraviolet and visible light. For designing any smart and self-reliant system various type of sensors are required as denoted in Fig. 1. Hence, these M2M devices can be used for many applications like smart cities, smart agriculture and factory, automobile parking, traffic management, water conservation, smart eco-friendly environment, and more. Considering, a water level sensor which is used for the early alarm of flood. This is implemented with the help of WSN in the areas that are severely affected by the floods [7]. To detect the water level, an ultrasonic sensor is required which is installed at a suitable location and the sensor measures the distance between the ground and water level as shown in Fig. 2. The sensor node is connected to the microcontroller and as the water level increases the light will glow and hence the buzzer sounds. Then, this signal through the Internet and radio transmission is sent to the monitoring authorities where the information is collected and visualized and a warning alert is issued if required.

3 Energy Conservation For the efficient operation of any smart IoT system, a continuous source of power is required and most of the wireless sensor systems which are currently in practice make use of the secondary storage device battery. But it is very catastrophic to use

52

N. Singh et al.

a battery for their operation, as battery requires frequent charging and has a limited lifespan which interrupts network consistent operation. Therefore, for the healthy operation of the IoT systems a lifelong, reliable and self-sustainable energy solution is needed [8–10]. A number of energy source existing in the surroundings are solar energy [11], thermal energy [12], vibrational energy [13, 14], acoustic energy [15], wind energy [16], tidal energy [17], and RF energy [18–20]. Among these stated energy sources, the RF energy is more advantageous as it is available in an optimal amount in the environment and does not require bulky conversion setup. The other energy sources depend upon the climatic or other environmental conditions and are difficult to harvest also. The various merits and demerits of different energy harvesting systems are given below.

3.1 Solar Energy Harvesting The solar energy harvesting system is the most commonly used method because of the easy accessibility of solar energy. The first study related to solar energy was demonstrated in 1950 on silicon-based solar cells and till date, the research is still continued for maximizing the conversion efficiency through miniaturized cells. Normally, the semiconductor diode based photovoltaic cells are used to convert light energy or solar energy into electrical energy; sometimes infrared or ultraviolet rays are also used for conversion. The materials used in the construction of photovoltaic cells are cadmium telluride (CdTe), gallium arsenide (GaAs) and copper indium diselenide (CIS). By connecting several photovoltaic cells in a series and parallel configuration, the highefficiency solar panels can be constructed. But the main drawback of this source of energy is the limited availability of sun in the daytime only. On an average, the sun is available for four to eight hours and based on the principle of photovoltaic effect the sun radiations can be converted into useable power. The other demerits of solar harvesting system are high installation and operating cost, low output efficiency, and bulkier panels.

3.2 Thermal Energy Harvesting Thermal energy can be directly converted into electrical energy through the principle of the Seebeck effect. When the heat is applied to two dissimilar metal conductors or semiconductors, a temperature difference is created, and this produces a voltage difference between the two materials. In the Seebeck effect or Thomson effect, the electrons from the heated conductor flow towards the cooler conductor and if the two pairs of metals are directly connected through some electrical circuit the voltage generated can be consumed directly. The direct current (DC) voltage generated by this

Energy-Efficient System Design for Internet …

53

system will be very small (in microvolts) and to increase the magnitude of generated voltage, an array can be used. These devices are called as thermoelectric devices and they can operate continuously without any delay till the time temperature difference is maintained. Unlike solar energy, thermal energy is available all the time but it needs a large area and the power produced is low. Another type, popularly known as Peltier effect, is present when two different conductors (for example A and B) show heating and cooling at an electrified junction, depending upon the direction of flow of current. This effect is stronger when two dissimilar semiconductors are used. When current flows through two different materials, based on the direction of current, the temperature of the junction rises or falls. For instance, when copper and bismuth wire are joined together in an electrical circuit, the heat is produced at the point where current passes from copper to bismuth and a fall in temperature in the reverse direction. The Peltier effect is reversible in nature and by changing the direction of current flow, the effect can be changed. This effect was discovered by the French physicist Jean Charles Athanase Peltier in the year 1834.

3.3 Vibrational Energy Harvesting This method converts mechanical energy into electrical energy by means of applying stress. “Piezo” is a Greek word, which means to press, push, or squeeze and the mechanical strains may be of any type either vibration or any deformation of material. The two scientists Jacques and Pierre Curie discovered this effect for the first time and demonstrated that by applying pressure on the quartz material, electric charge can be produced. In Fig. 3, two metal plates are used at the upper and lower side of the piezoceramic material also known as crystal and as the mechanical stress is applied across the material, the positive and negative charges are produced on the metal plates which are responsible for the generation of electric charge. The commonly used devices that utilize the piezoelectric effect are microphone, electric guitar, loudspeaker, hydrophone, pressure sensors, etc. The most common voice recognition application “Siri” used in smart devices makes use of the piezoelectric effect when a person says something in the smartphone microphone, the piezoelectric material converts the sound signal vibrations into electrical signal to accomplish the given voice command. Both natural and manmade materials can be used for the generation of piezoelectric effect. Some of the well-known natural materials are berlinite, Rochelle salt, tourmaline, cane sugar, and topaz. Few manmade materials are barium titanate and lead zirconate titanate. The vibrational energy harvesting systems are light in weight but they require a large area for their operation and also the produced output is not fixed.

54

N. Singh et al.

Fig. 3 Piezoelectric plates used for generating electric charge

Fig. 4 Block diagram of electrostatic energy harvesting system

3.4 Electrostatic Energy Harvesting The electrostatic converters are mainly consisting of an energy transfer circuit and variable capacitors. The two plates of the variable capacitor are kept at some distance and mechanical stress is applied at one end of the capacitor plate, which changes the value of the capacitor and this variation in capacitance value results in the change in voltage or current. Figure 4 represents the block diagram of the electrostatic energy harvesting system. The output from the energy transfer circuit can be utilized directly by a load resistor or it can be stored in the form of chemical energy also. The electrostatic energy harvester circuit is simple, compact, easy to design and integrate with microelectronics, compatible with microelectromechanical (MEMS) systems but the only limitation is the small quantity of power generated. A comparison of five different types of variable capacitors with their characterization is shown in Table 1. Depending upon the input power requirement and application of the structure, a particular type of varicap circuit can be chosen.

3.5 Wind Energy Harvesting The wind power can be harvested with the help of large wind turbines by mechanical means and converted into electrical energy. Basically, two types of wind conversion

Energy-Efficient System Design for Internet …

55

Table 1 Comparison of different types of variable capacitor References

Type of variable capacitor

Characterization

[21]

In plane overlap variable capacitor

Needs power processing

[22]

Varying capacitance machine

Can be used for HVDC system and capacitance value is directly proportional to the number of poles

[23]

Tunable capacitor

Glass substrate is used to minimize the parasitic capacitance

[24]

In plane gap closing

Relative displacement amplitude is kept smaller than the gap of two fingers

[25]

Electrostatic swing harvester

Generates DC output voltage around 3V

systems are used for wind energy harvesting; one is fixed speed type and another is variable speed type. The fixed speed system is directly connected to the grid whereas the variable speed system works as an electronic interface. The wind turbine is the main unit which converts wind power into mechanical energy. The wind turbines are categorized as the vertical axis and horizontal axis type. Vertical Axis Wind Turbine (a) The main rotor shaft is kept perpendicular or in the vertical position with respect to the ground. (b) The replacement and repair of generator and gearbox are easy as these components are near to the ground. (c) The blades of these turbines need not be pointed as it will receive wind from all the directions. Horizontal Axis Wind Turbine (a) The main rotor shaft is kept in parallel or in a horizontal position with respect to the ground. (b) The most commonly used horizontal axis wind turbines comprised of two or three blades. (c) When the wind blows a low-pressure air is produced at the back side of the blade and high-pressure air in the front side of the blades. (d) A rotor shaft is connected to the generator and this electrical generator harvests the wind energy by converting into mechanical and then to DC current.

3.6 RF Energy Harvesting The RF energy conversion also known as wireless energy harvesting (WEH) can be realized with the help of a device called rectenna [26, 27]. The rectenna senses the electromagnetic waves available in the environment with the help of a compact

56

N. Singh et al.

Fig. 5 Different types of energy harvesting systems

electronic circuitry. Rectenna is basically comprised of an antenna which senses the electromagnetic signal, rectifying circuit for converting RF signal to DC waveform, a matching circuit to match the impedance of the antenna and rectifying diode, and a DC pass filter for smoothening and removing unwanted ripples from the output waveform. Hence, the rectenna is an emerging easy access energy solution to M2M communication for IoT devices. Figure 5 represents the different types of energy harvesting systems which can be employed for powering up smart IoT devices. The rectenna used for RF energy harvesting was discovered in early 1990 by W.C. Brown. At present, the major design challenge faced by electrical engineers is the low RF to DC conversion efficiency of the rectenna. Several designs have been developed to increase the output intensity of the rectenna so that several IoT devices can be power up simultaneously. Various techniques like an array of antenna, voltage doubler rectifier, Villard rectifier, use of numerous frequency bands, multiband, and

Energy-Efficient System Design for Internet …

57

Table 2 Merits and demerits of energy harvesting systems Type

Source

Power density

Advantage

Disadvantage

Solar energy

Sunlight

mW/cm2

Limitless availability

Not available during night time

Wind energy

Wind

mW/cm2

High power received

Availability depends on weather conditions

Piezoelectric energy

Vibration

μW/cm2

Both indoor and outdoor applications

Small life span of device

Thermal energy

Heat

μW/cm2

Easy to implement

Less power density

RF energy

EM waves

μW/cm2

Freely available during day and night

Distance is the limitation

broadband antenna that can utilize the ambient energy available in the environment have been developed. The use of circularly polarized sensing antenna might be a better approach, to increase the conversion efficiency, since it suppresses the multipath fading effects present between sensing element and transmitter. Moreover, several types of circularly polarized rectenna have been developed for biomedical and smart sensing devices. The performance comparison of different energy harvesting systems is shown in Table 2.

4 Harvesting Module A block diagram of the energy harvesting module is shown in Fig. 6. It is mainly consisting of three blocks: a rectenna for RF-DC conversion, DC-DC converter and a power management unit (PMU). The converted energy can be used to power up the sensing node directly or the energy may be stored for future use. For the proper functioning of the IoT system, the sensing node directly utilizes the harvested energy and when this consumption is less than the total amount of harvested energy, the storage component stores the excess quantity. The various blocks of energy harvesting system are explained in the next subsections.

4.1 Rectenna Model Figure 7 represents the block diagram of the rectifying antenna popularly known as rectenna. The first block is a receiving antenna which senses the electromagnetic waves from the surroundings. Thereafter, a matching circuit is present which matches

58

N. Singh et al.

Fig. 6 Block diagram of wireless powered IoT device

Fig. 7 Block diagram of the rectenna

the antenna impedance with the impedance of the rectifying diodes. To obtain maximum RF to DC conversion efficiency from the rectenna, the matching circuit needs to be designed carefully. Further, the unwanted harmonics generated due to the nonlinear behavior of the rectifying diodes will be blocked by the DC filter shown in the last block.

4.2 Sensing Antenna The different types of sensing antenna can be monopole [28], dipole [29], Yagi–Uda [30], folded dipole [31], fractal [32], bowtie [33], slot [34], and microstrip patch antenna [35, 36]. For efficient wireless energy harvesting, the antenna must have high gain and multiple radiating band characteristics. The different types of antenna and their performance characteristic are explained below. Bowtie Antenna It is a subpart of the biconical antenna also called as butterfly antenna. Figure 8 shows the planar biconical antenna which has the advantage of wide bandwidth, low profile, and also acts as a high pass filter in the rectenna systems. The bowtie antenna shows a higher gain at low frequency. In the figure, it can be seen that a feeding line is present in the center of bowtie to excite the antenna. This bowtie antenna shows similar radiation behavior as of dipole antenna.

Energy-Efficient System Design for Internet …

59

Fig. 8 Layout of the bowtie antenna

Fig. 9 Layout of the log periodic antenna

Log Periodic Dipole Antenna It is an array of dipoles with variation in size and the last dipole is considered as the largest element whereas the first dipole at the front end is the smallest element. The antenna elements are connected in parallel to each other and with feedline and it is very commonly used for the purpose of broadcasting. Each element of the array is connected with the alternating phase. The log periodic dipole antenna array is a high gain wideband antenna operating in the ultra high-frequency region. The Log periodic antenna design is shown in Fig. 9. Wire Antenna The wire antenna or short dipole antenna where the term short refers to the size and the size of wire antenna is completely dependent upon the wavelength. A short dipole antenna is a very common and simple type antenna with feeding line connected in the center and its length should be less than the one-tenth of the wavelength. This antenna is omnidirectional with 90° half-power beamwidth and shows a linear polarization

60 Fig. 10 Layout of the diopole antenna

N. Singh et al.

Quarter-wavelength

Feeder

radiation characteristic. This antenna is mostly used for narrowband transmission applications. Dipole Antenna A dipole antenna is an extension of the short dipole and in this antenna, the length of each conductor must be chosen as the quarter wavelength of the operating frequency. The dipole antenna is simple to design and easy to fabricate. Compared to short dipole antenna, the dipole antenna is directional and has a large bandwidth. Figure 10 represents the layout of a popularly used dipole antenna. Monopole Antenna As the name monopole suggests these antennas are with single pole and consists of a single conductor generally connected in the perpendicular or vertical direction. The antenna size is the half wavelength of the simple dipole antenna. The directivity of the monopole antenna is roughly double the length of the antenna. This is because of the antenna radiation which is mostly above the ground plane. This antenna was discovered by Guglielmo Marconi and for this reason, monopole antenna is also called as Marconi antenna. A simple monopole antenna is shown in Fig. 11. Loop Antenna The loop antenna can have a circular, rectangular, ellipse, pentagon, or hexagon geometry. The circumference of the loop determines the antenna efficiency similar to monopole where the conductor length measures the efficiency of the antenna. On the basis of circumference, the loop antenna is classified as of two types; one is electrically small whose circumference P j,curr ent

= 0 and the lower bound of the total servers is calculated as [ The inequality is given by

m j=1

P j,curr ent P j,max

]

Models and Algorithms for Energy Conservation in Internet of Things m 

 m ej ≥

j=1

j=1

P j,curr ent

P j,max

91

 (2)

The exact and extended bin can be e j = 1 elso 0 if the server is not used and xi j = 1 if V Mi is placed in server j otherwise 0 [48, 50].

6.2 Best Fit Heuristic Algorithm This is used to save energy and it is of two steps. It sorts the requested VMs from higher to lower order of energy consumption to build the ordered stack. The next step involves, the VMs are always handled from the top of the stack. The most energy consuming servers are packed with least energy until the remaining servers of VM down the stack fits the target server. This process is repeated until all the VMs are packed with the most occupied servers. This will allow the servers to enter into sleep mode or switch off when they are in free or idle state [5, 51].

7 Dynamic Energy-Efficient Algorithms The current utilisation of the energy by the physical machine is treated as dynamic energy consumption. It depends on the I/O activity, clock rates and the CPU usage scenario. Dynamic energy can be consumed in two ways. Switched capacitance which performs charging and discharging of a capacitor. Short circuit current is a minor source of dynamic energy consumption. The dynamic energy consumption can be reduced by reducing the switching activity and reduce the physical capacitance, reduce supply voltage, reduce the clock frequency. The dynamic energy efficient techniques are further classified into Hardwarebased, energy-efficient algorithms and software-based, energy-efficient algorithms.

7.1 Hardware Level Solution The hardware energy efficient technique works in fulfilling the requested services using minimum number of components. This will further switch off the idle components of the system, varying the frequencies from minimum to maximum. Benini designed a method of switching the frequency based on the CPU mode of operation and I/O mode of operation. DVFS algorithm for the hardware level solution is discussed in the further section.

92

7.1.1

B. Rudra and D. S. Prashanth

Dynamic Voltage Frequency Scaling (DVFS)

Dynamic Voltage Frequency Scaling (DVFS) [52] strategy is used for reducing the power consumption by modifying the CPU frequency according to workload. Voltage scaling is to reduce the energy consumption by adjusting the voltage to a threshold value without degrading the output. It scales the voltage and frequency of the CPU during the execution process. It is mainly used to enhance the energy-efficient scheduling on the servers that are in idle state, under light workloads during the execution of noncritical tasks. This can be activated in four modes like High-frequency mode, Low-frequency mode, available frequency mode and on-demand frequency mode. The requested services arrived at the cloud may be a CPU bound request or an I/O bound or Memory bound request. The incoming packet follows any algorithm to reach the cloud. DVFS [53] depends on hardware and not changeable according to the needs that are varying. The consumption of the power with this method is not that much reduced compared with the available other methods. As it works at server level, switching of servers when not in use will reduce the energy consumption. Many servers in the centers will be in idle mode at many times, can be made to sleep mode or power down at particular periods of time and made power up when in use. The key method is to reduce the load and distribute the load to small number of VM servers which aims to reduce high consumption of energy. Using VM live migration the reduction of energy consumption can be achieved [5].

7.2 Software Level Solution The DVFS-based hardware-level solution will reduce the power consumption of PM by switching-off the PM present at idle condition [54]. Thus, there is a requirement to find a solution to avoid the switch-off condition which is present in hardware solution. Dynamic power consumption at software level of cloud data center can be reduced by using some energy-efficient AM allocation and migration algorithms, Task scheduling algorithms, network aware VM allocation algorithm, etc. The Bioinspired algorithms for the migration of VM to PM in order to reduce the energy consumption are emerging and reaching towards success. The details of some of the algorithms are discussed in the following sections.

7.2.1

First Fit Decreasing Algorithm (FFD)

In this algorithm, first the bins are arranged in decreasing order, select the largest bin and examine to fit the VM in the package considering the size and packing cost. The result has shown that the implementation of FFD proved to save 25% of energy [51, 55, 56]. The algorithm steps are as follows:

Models and Algorithms for Energy Conservation in Internet of Things

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

15.

93

List out all the Individual Tasks with its respective deadlines Set a value by considering the maximum utilisation rate For all the requested task (new) Arrange all task from the higher order to the lower one based on the Deadline and Resource requirement Calculate end time of the respective task Arrange all the nodes of the cloud in decreasing order based on their utilisation rate. Arrange all the VMs in decreasing order based on the resource usage. Set value for finding to reuse the VM is false Check whether the physical elements (Each physical element (PE) to run a set of VM) are lesser than the threshold for the PE. If a VM is allocated, change the value for finding to reuse the VM is true, list a task in VM If the rate of utilisation is greater than the maximum utilisation rate then assign utilisation rate value to the maximum utilisation rate If VM which is not able to reuse is false then initialize a new VM and allocate a task, repeat the above step For a group of Physical Elements, a node is connected For every virtual machine in every physical element, If VM CPU is lesser than the minimum threshold, then identify the VM with less load where the load can be migrated to the other VM and shut down the previous VM If PE CPU is greater than the maximum threshold then identify lesser loaded PE using greedy algorithm.

7.2.2

Modified Best Fit Decreasing Algorithm (MBFD)

Beloglazov et al. modified the Best Fit Decrease (BFD) algorithm and suggested a modified algorithm. In this Algorithm of Modified Best Fit Decreasing (MBFD) [57, 58], the VMs are sorted from higher to lower order based on the current utilisation of CPU. Allocate each VM to a PM which decrease the energy consumption. It is given by x*y where x are the number of VMs and y are the number of hosts. The steps are as follows: 1. Make a list of all the hosts and VMs on the cloud server with all the available resources. 2. Arrange all the n VMs from higher order to lower order based on their utilisations 3. Consider each VM in list and check out the VM which requires minimum power to maximum power and list the m Hosts which are NULL. 4. List all the available host 5. If VM meets the host requirements, then estimate power consumption by both host and VM Else go to the step 4

94

B. Rudra and D. S. Prashanth

6. If power requirement is lesser than the lowest power, then host is allocated to a VM with minimum power Else go to the step 4 7. If Hosts list has a NULL host then map a VM to the Host i.e n to m Else go to the step 3

7.2.3

Genetic Algorithm (GA)

Genetic Algorithm (GA) is an adaptive heuristic search algorithm based on natural selection and genetics. This algorithm is used as optimisation technique for the conservation of energy. The genetics are selected randomly, but no means of random, as it is based on the historical informaton. The historical information is obtained by performing the search in a region for better results in a space. The individual population is maintained in the search space of GA [59] by presenting a possible solution. It is designed to perform simulation processes in natural systems for the purpose of evolution [60]. Each individual in the population is coded with a finite vector variables. Each individual is called as chromosomes and its variables are the genes. Each Chromosome consists of several variable functions. A fitness value is assigned to chromosomes. Sorting operation is performed to find the optimal solution or near by optimal solution. Once the initial population is generated, the algorithm performs the following operations. • Selection Operator: It is used to select the best individuals from the set of population available and used to generate parents. Generate fitness value, sort them according to their decreasing value. Based on the fitness value, preference to the individuals are given allowing them to pass the genes to the next level. Sometimes the usage of roulette wheel may loose the excellent chromosomes due to random errors. Calculate the average fitness value, divide the group into two parts based on the average value where one part p1 consists the individuals whose fitness value if greater than the average and the other part p2 consists of less than the average value. Select each from one part using the selection operator, where p1 member ensures better value and p2 member has less fitness which is used to maintain the population diversity. • Crossover Operator: It is used for choosing the matching chromosomes and combining them to generate next level of population. The chromosomes are chosen from the two parts of population using the selection operator. • Mutation: Performs random changes. Some of the bits with low probability, which are flipped. This is used to maintain diversity in the population. The modified Genetic algorithm [60] is applied for the energy efficiency in cloud. It consists of four steps like Initialization, evaluation, exploitation and exploration. For each iteration, it selects a VM at random from the center and replace them at optimal place. It selects the VM with high configuration by appling selection rules, The combination of VMs is done based on capacity of size, applications, and the optimal solution for next generation. Perform random changes on the properties to

Models and Algorithms for Energy Conservation in Internet of Things

95

create virtual machines. Intialisation is performed by randomly collecting all the local VMs load and the Original VMs placement details. Evaluate the fitness of all the VMs in the iteration. Execute the modified Genetic algorithm which consists of selection, crossover and mutation process. Compare with the collected original VM placement, if the result is 1 then perform migration of the VMs else collect the Vm details and perform the GA operations. Once the migration is performed go to next iteration. Repeat the same procedure for all the chromosomes or VMs present in the iteration to find the optimal solution [61].

7.2.4

Particle Swarm Optimization (PSO)

Particle Swarm Optimization (PSO) is used to find the fitness of each individual particle leads to a feasible solution to the problem present in the cloud. Each particle consists of velocity and position [6, 62]. The quality is measured based on the fitness value associated with the particle position. The algorithm starts with the initialization of a group of particles randomly, finding the optimal solution by performing iterations. Every particle in the group changes its position with a specified velocity by searching the local best position X lbest,i and the global best position X gbest,i . The velocity of each particle along with the positions are updated by the following equations. Vit+1 = pVit + c1r1 (X lbest,i (t) − X it ) + c2 r2 (X gbesti (t) − X it )

(3)

X it+1 = X it + Vit=1

(4)

Vit and Vit+1 are the velocity before updated and after updated respectively, X it+1 are the before and after updated position respectively. The p is the inertia weight coefficient which defines the current inheritance velocity of the particle, balanced by local and global search capabilities. The learning factors are given by c1 and c2, and the random numbers lies between 0 and 1 are presented by r1 and r2. Particle Position It is defined by n bit vector where n is the particle code which is equal to the number of servers in the data center. The position is given by X it = t t t , xi2 ....xin ). (xi1 t t t , vi2 ....vin ) where n is the bit vector that Particle Velocity It is defined by Vit = (vi1 t presents the adjustment decision of the VM. Vi guides the particle position which is 0 or 1. If the re-evaluation and the adjustment is required then 0 otherwise 1. Subtraction Operator It calculates the distance between two VMs and represented by Θ. It is called as binary subtraction operator in PSO. If the operands contain same bit value then the position of the vector value is 1 else 0. Addition Operation In PSO we have to consider local optima solution as well as Global optima solution of the fitness value. In order to find the best fitness value, we require global optima solution and local optima solution. For this, we consider three inertias along with uncertainty bit q. The final fitness value along with the inertia is as follows. X it ,

96

B. Rudra and D. S. Prashanth

f (X it )

=

m 



j=1

(P j 2 ) + (u dj − 1)2 P j max

(5)

The inertia equations are as follows. P1i =

P2i =

P3i =

f (X it )

f (X it ) t t + f (X lbest,i ) + f (X gbest ) t f (X lbest,i )

t t f (X it ) + f (X lbest,i ) + f (X gbest ) t ) f (X gbest,i t t f (X it ) + f (X lbest,i ) + f (X gbest )

(6)

(7)

(8)

t t where f(X it ) fitness value of ith particle of the solution. X it , X lbest,i and X gbest,i are the local, global best position of theith particle. In order to calculate the inertia, Consider the high-energy efficieny amd maximum utilisation with high probability. we consider three uncertain bits as there are three inertias. These bits are given by uncer tainbit = q1i f rand ≤ P1i , uncer tainbit = q2i f P1i < rand ≤ P2i and uncer tainbitq3i f P2i < rand ≤ P3i . q1 is the bit value of particle before updating, q2 is the bit value of local best particle and q3 is the bit value of global particle.  Multiplication Vector It is represented as and used to update the position. t i and the vector velocity as Vit=1 then the If the current position s given by X P  t+1 t Vi . The rule of multiplication is, if the bit value of multiplication works as X i the velocity vector is 1 then the corresponding position vector is not disturbed else it will be adjusted to the required position. Virtual Machine Placement Optimization First the request is collected from VM on a periodical basis is treated as the input. Based on the VM request and the available resources, the initial population is going to be generated. The First fit strategy is adopted to generate a set of population with feasible solution. The initial particle position is found by initial population. The initial velocity is determined based on the status information in first dimension. Once the fitness value is obtained, the local best position and global best position is found. Update the particle velocity using the following equation

Vit+1 = P1 V1t



P2 (X lbest,i (t)Θ X it )



P3 (X gbest (t)Θ X it ).

(9)

 t+1 Update the positions of the particles using X it+1 = X it Vi . This sometimes may be in loss of losing some of the virtual machines which are refilled back.

Models and Algorithms for Energy Conservation in Internet of Things

97

This operation may lead to the duplicate virtual machines which are to be removed to ensure the feasibility. Sometimes this removal may lead to the removal of virtual machines that are in non-repetitive mode. However, Reallocation of the virtual machines will solve the problem by reinserting the VMs to the servers. This reinserting is called reallocation. In the reallocation policy, it follows First Fit algorithm along with consideration of active servers as host servers. A new server is turned on when the current server is unable to place the VM. The local and global positions are updated based on the updated new population of particles. The process repeats till it finds the current iteration value as greater than the specified maximum iteration value. The process ends up by finding the global best position along with its fitness value. Once the optimal solution is found, the VMs request is placed on current server and the process ends.

7.2.5

Ant Colony Optimization (ACO)

The requests from the users contain the details of the resources required. All the requests are kept in the queue and maintained by the controller. Ant Colony Optimization algorithm helps to allocate and maintain the cloud services in systematic way. 1. Initially, listing out all the available valid nodes and a pointer pointing to it. 2. If a request from a customer receives: 3. If the requirements of a customer match with the available node, then it creates a basic Virtual Machine. 4. Create a Monitor ant to supervise the Virtual Machine for balancing the load. It always returns three values one for mentioning balancing is not required, other mentions it is recommended for balancing and the final value is strictly balancing is required. Go to step 7. 5. If the capacity of the pointed node does not meet the requirements of the customer then automatically pointer moves to next available valid node. Go to step 4. 6. If all the nodes in the list doesnt happen to meet the requirements then alert the admin mention the shortage. 7. Monitor ant will return Throughput and Response Time Throughput THRPUT, Response Time RT and Average Response Time AvgRT 8. If RT < 0.9 * AvgRT AND THRPUT > 1.1 * THRPUT then return 0 9. Else if 0.9 * AvgRT< RT < 0.95 * RT (OR) 1.10 * THRPUT > THRPUT > 1.05 * THRPUT Then if both the conditions are true then the VM is suggested for balancing and VM is clonned Otherwise if only first condition is satisfied then also VM is suggested for balancing the Load and VM is cloned Else if only second condition is satisfied then VM is suggested for balancing the Load gets migrated 10. Else if RT > 0.95 * AvgRT (OR) THRPUT < 1.05 * THRPUT then Then if both the conditions are true then the VM is severely need of balancing the Load and VM is cloned Otherwise if only first condition is satisfied then also VM is severely

98

11.

12.

13.

14.

15.

B. Rudra and D. S. Prashanth

need of balancing the Load and VM is cloned Else if only second condition is satisfied then VM is severely need of balancing the Load gets migrated Testing Ant: If processor utilisation (or) memory consumption is greater than 90% then arrange the VMs from higher to lower order and select the top most VM If VM Monitor returns value = 2 then go to step 11 If VM Monitor returns value = 1 then go to step 13 If processor utilisation (or) memory consumption is less than 50% then consider all the other nodes capacity and arrange them from the higher capacity to the lower. If there is possibility to migrate all the VMs to a single node then migrate them. If the monitor mentions that node is critically in need of Load Balancing and clone the VM and if any node is available with enough resources allot the respective node. If there are VMs whose monitor returns that any node has more than 30% of resources available then active the power effective node. If the particular node is standby activating the selected node If all the nodes are utilised in the list then alert the admin mention the shortage.

S is used to count ants which are used further to discover the latest cloud providing the services and adds to the resource table. This can be achieved by registration of node.Cleaner ant updates the resource table by creating an agent, it goes into the cloud and wait for the response for a particular duration. If it fails to get response, agent considers that cloud is failed [63–66].

7.2.6

Simulated Annealing (SA)

The basic idea is taken from annealing process which has three stages [67–69]. Heating stage is used to enhance the thermal motion of particles to deviate from its position. Isothermal Phase for the exchange of heat with surrounding within a closed room of constant temperature. The system changes its temperature spontaneously reducing the energy. When the energy is reduced to a minimum, the system is in equilibrium. Cooling Stage is for making the thermal particle weaken where the energy decreases gradually and form it into a low energy crystal structure. It is applied to reduce the workload of VMs by using task scheduling which is as follows. Select the initial value x and fix a high temperature. The task scheduling is given by E = f (x) where E is the energy and F(x) is the general function. For initial task scheduling, it uses Greedy method which uses the task priority principle. Select three VMs say as m,n,o and set them to PMs M,N,O. Resecure these as m to N physical machine (PM), n to K Physical Machine (PM), O to M Physical machine (PM). calculate new values E new = f (xnew ). If E new < E, then accept the new path else accept it after some probability of time and cool down the system. If path is accepted then stop else repeat the process of assigning VMs to PMs. If E(y) is the function value at y and Y is the current state and

Models and Algorithms for Energy Conservation in Internet of Things

99

Y* is the new state, r is to control the speed, T is temperature and T ∗ is the minimum temperature. If T is equal to T* then stop the process the cooling of the center is done else follow the following steps. W hile(T > T ∗ )then E = E(Y ∗ ) − E(Y ) i f (E ≥ 0)Y ∗ = Y happens only when more optimal solutions are found else i f (ex p(E/C T )) > random(0, 1)wher eCisaconstantsystem.T henY ∗ = Y and T = r ∗ T . If 0