Molecular Communications: An Analysis from Networking Theories Perspective 3031368819, 9783031368813

Table of contents :
Preface
Technical Supervisors
Contents
List of Figures
List of Tables
1 Introduction
References
2 Analysis of Layer's Tasks in Molecular Communication: Application, Transport, Network, and Link Layers
2.1 Molecular Communication Network Architecture
2.1.1 Molecular Application Layer
2.1.1.1 Molecular Application Layer Functionalities
2.1.2 Molecular Transport Layer
2.1.2.1 Molecular Transport Layer Communication Model
2.1.2.2 Molecular Transport Layer Functionalities
2.1.2.3 Significant Cases of Investigations in Molecular Transport Layer
2.1.3 Molecular Network Layer
2.1.3.1 Molecular Network Layer Communication Model
2.1.3.2 Molecular Network Layer Functionalities
2.1.3.3 Significant Research Studies in Molecular Network Layer
2.1.4 Molecular Link Layer
2.1.4.1 Molecular Link Layer Communication Model
2.1.4.2 Molecular Link Layer Functionalities
2.1.4.3 Significant Cases of Investigations in Molecular Link Layer
References
3 Analysis of the Molecular Physical Layer's Tasks
3.1 Definition of Nanomachine and Bio-nanomachine
3.2 Definition of Nanonetwork (Bio-nanonetwork)
3.3 Molecular Communication Systems
3.3.1 Molecular Communication System Components
3.3.2 Theoretical Modeling of Molecular Communication
3.3.2.1 Random Walk
3.3.2.2 Random Walk with Drift
3.3.2.3 Random Walk with Reaction by Amplifiers
3.3.3 ISI, Noise Analysis, and Error Control Techniques
3.3.3.1 Molecular Communications and the Impact of Noises
3.3.3.2 A Classification of Noises and Their Sources
3.3.3.3 Error Control Techniques
3.3.4 Analysis of Physical Estimation Parameters
3.3.5 Molecular Modulation
3.3.5.1 Time Modulation
3.3.5.2 Concentration-Based Techniques
3.3.5.3 Type-Based Techniques
3.3.5.4 Timing-Based Techniques
3.3.5.5 A Hybrid Type of Modulation
3.3.6 Molecular Physical Layer
3.3.6.1 Bio-nanomachine Sublayer
3.3.6.2 Signaling Sublayer
References
4 Case Studies of Applications of Digital Networks Theories to Molecular Network Stacks
4.1 Case 1. Bacterial Molecular Communication BasedNanonetworks
4.2 Case 2. Internet-of-Nano-Things Healthcare Applications
4.3 Case 3. Modeling Nonviral Gene Delivery as a Macro-to-Nano-Communication System
4.4 Case 4. Internet of Things for Advanced Targeted Nanomedical Applications
4.5 Case 5. Hybrid DNA- and Enzyme-Based Computing for Address Encoding, Link Switching, and Error Correction in Molecular Communication
4.6 Case 6. Efficient Molecular Communication Protocol Based on Mobile Ad Hoc Nanonetwork
4.7 Case 7. IEEE Standard Data Model for Nanoscale Communication Systems
4.8 Case 8. Gene Expression and Protein Delivery Analysis from a Network-Layered Paradigm
References
Index

Citation preview

Yesenia Cevallos · Cristian Vacacela Gómez · Luis Tello-Oquendo · Talia Tene · Deysi Inca · Ivone Santillán · Albert Espinal · Nicolay Samaniego

Molecular Communications An Analysis from Networking Theories Perspective

Molecular Communications

Yesenia Cevallos • Cristian Vacacela Gómez • Luis Tello-Oquendo • Talia Tene • Deysi Inca • Ivone Santillán • Albert Espinal • Nicolay Samaniego

Molecular Communications An Analysis from Networking Theories Perspective

Yesenia Cevallos College of Engineeering Universidad Nacional de Chimborazo Riobamba, Ecuador Universidad San Francisco de Quito IMNE, Diego de Robles s/n Cumbayá Quito, Ecuador Luis Tello-Oquendo College of Engineeering Universidad Nacional de Chimborazo Riobamba, Ecuador Deysi Inca College of Engineeering Universidad Nacional de Chimborazo Riobamba, Ecuador Albert Espinal Faculty of Electrical and Computer Engineering Escuela Superior Politécnica del Litoral Guayaquil, Ecuador

Cristian Vacacela Gómez Escuela Superior Politécnica de Chimborazo Riobamba, Ecuador Talia Tene Universidad Técnica Particular de Loja Loja, Ecuador Ivone Santillán College of Health Sciences Universidad Nacional de Chimborazo Riobamba, Ecuador Nicolay Samaniego College of Education, Human Science and Technologies Universidad Nacional del Chimborazo Riobamba, Ecuador

ISBN 978-3-031-36881-3 ISBN 978-3-031-36882-0 https://doi.org/10.1007/978-3-031-36882-0

(eBook)

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 This work is subject to copyright. All rights are solely and exclusively licensed 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 Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland Paper in this product is recyclable.

Preface

Technological advances in engineering and biology have allowed us to join efforts and produce efficient models using mimics of nature and its characteristics. Industry, medicine, engineering, biochemistry, biotechnology, computer sciences, and other disciplines apply these models. Bionanotechnology is a branch of nanotechnology that employs biological materials for medicine or biotechnology applications. In that order, molecular communication (MC) has attracted a significant attention in bionanotechnology as a modern technique that uses biochemical signals to achieve the interchange of information among components naturally and artificially, creating bio-nanoscale machines throughout short distances. The demand for genomic signal processing is growing drastically because of the critical role in the treatment of diseases and due to the importance of human genetics and related sciences. Essentially, molecular communications (MCs) analyze the information conveyed at the nano level. The cells, smart devices that constitute our bodies, mainly communicate fundamentally through the transport and binding of molecules. Opposite to existing telecommunication paradigms, MCs use molecules as information carriers. Sender biological nanomachines (bio-nano machines) encode data on molecules (signal molecules) and release the molecules into the environment. The molecules then travel through the medium to reach the bio-nano machine receptors for a biochemical reaction with the molecules to decipher the information. MCs have several applications, such as the diagnosis and treatment of diseases. The emulation of nanoscale biological processes allows for personalized predictions of disease evolution. Moreover, the design of nanosensors/actuators could detect and treat a large set of diseases (e.g., cardiovascular diseases and tumors). These nanodevices could activate the immune system or trigger drug delivery systems in small specific areas without affecting the rest of the body (i.e., smart drugs). We can remark that other fields, such as food production, functionalized materials, fabrics, and environmental/military applications, leverage MC systems. Communication frameworks in digital networks, such as the Open Systems Interconnection model (OSI) and Transmission Control Protocol/Internet Protocol (TCP/IP), have already been considered for MC system design. Thus, computer v

vi

Preface

network researchers visualize MC as a new communication paradigm, turning it into a recognized discipline in computer networks in which the models and architectures support models and tools (e.g., information theory) for MC research. Similarly to networked computing devices, biological cells can transmit, receive, and process information. Just as data communication protocols resulted in the rapid growth and ubiquity of networked computing devices and applications, the development of communication protocols for nano-based networks will stimulate groundbreaking future applications of bio-nanodevices. MC architectures can approach computer network perspectives through the TCP/IP reference model. Therefore, it can assess how information flows from a source through a router to the destination, representing clusters of bio-nanomachine senders, bio-nanomachines with routing features, and bio-nanomachine receivers, respectively. The source and router, as well as the router and destination, are within a communication range, which means that molecules’ information can propagate from one to another within a reasonable timeline to induce the intended reactions. Like the TCP/IP reference model, the application layer provides a set of options to implement applications. The network and link layers provide mechanisms to transmit information through and within a communication range. As a result, the network layer at the source selects a communication channel to convey the information to the link layer and guarantee channel availability. The physical layer contributes to biophysical mechanisms for the transmission and reception of information molecules over physical media. Hence, the physical layer transmits and propagates molecules’ information through the medium to the router. Then, the router selects a communication channel to meet its availability for molecules’ information-type propagation to the destination. The importance of MCs from a digital network perspective can be seen in the fundamental task of the data link layer to transform an imperfect channel into an error-free transmission or report unsolved problems to the upper layer. This approach applied to MCs could play a critical role in facilitating health applications (among others) development; for example, in targeted drug delivery systems, it can be helpful to increase the effectiveness in cancer treatments. Then, in this book, we analyze MCs from a networking perspective. Riobamba, Ecuador Riobamba, Ecuador Riobamba, Ecuador Loja, Ecuador Riobamba, Ecuador Riobamba, Ecuador Guayaquil, Ecuador Riobamba, Ecuador

Yesenia Cevallos Cristian Vacacela Gómez Luis Tello-Oquendo Talia Tene Deysi Inca Ivone Santillán Albert Espinal Nicolay Samaniego

Technical Supervisors

Jesús B. Alonso Department of Signaling and Communications, University of Las Palmas de Gran Canaria, Spain Jennyfer Granizo Metropolitano Hospital, Ecuador Floriano De Rango Department of Computer Engineering, Modeling, Electronics and Systems, University of Calabria, Italia

vii

Contents

1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 9

2

Analysis of Layer’s Tasks in Molecular Communication: Application, Transport, Network, and Link Layers. . . . . . . . . . . . . . . . . . . . . . 2.1 Molecular Communication Network Architecture . . . . . . . . . . . . . . . . . . . . 2.1.1 Molecular Application Layer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.2 Molecular Transport Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.3 Molecular Network Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.4 Molecular Link Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

13 14 18 20 26 45 56

3

Analysis of the Molecular Physical Layer’s Tasks. . . . . . . . . . . . . . . . . . . . . . . . 63 3.1 Definition of Nanomachine and Bio-nanomachine . . . . . . . . . . . . . . . . . . . . 66 3.2 Definition of Nanonetwork (Bio-nanonetwork) . . . . . . . . . . . . . . . . . . . . . . . 68 3.3 Molecular Communication Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 3.3.1 Molecular Communication System Components . . . . . . . . . . . . . 71 3.3.2 Theoretical Modeling of Molecular Communication . . . . . . . . . 81 3.3.3 ISI, Noise Analysis, and Error Control Techniques . . . . . . . . . . . 86 3.3.4 Analysis of Physical Estimation Parameters . . . . . . . . . . . . . . . . . . 95 3.3.5 Molecular Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 3.3.6 Molecular Physical Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

4

Case Studies of Applications of Digital Networks Theories to Molecular Network Stacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Case 1. Bacterial Molecular Communication Based Nanonetworks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Case 2. Internet-of-Nano-Things Healthcare Applications . . . . . . . . . . . 4.3 Case 3. Modeling Nonviral Gene Delivery as a Macro-to-Nano-Communication System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

167 168 169 171

ix

x

Contents

4.4 Case 4. Internet of Things for Advanced Targeted Nanomedical Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Case 5. Hybrid DNA- and Enzyme-Based Computing for Address Encoding, Link Switching, and Error Correction in Molecular Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6 Case 6. Efficient Molecular Communication Protocol Based on Mobile Ad Hoc Nanonetwork. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7 Case 7. IEEE Standard Data Model for Nanoscale Communication Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.8 Case 8. Gene Expression and Protein Delivery Analysis from a Network-Layered Paradigm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

175

177 178 180 184 192

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

List of Figures

Fig. 1.1 Fig. 1.2 Fig. 2.1 Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig.

2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10

Fig. 2.11 Fig. 2.12 Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig.

2.13 3.1 3.2 3.3 3.4 3.5 3.6 3.7

Two main directions for applying molecular communication theory to human health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Information theory paradigm to analyze molecular communications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The flow of information through a molecular communication network is modeled after the WAN architecture . . A layered architecture of molecular communication . . . . . . . . . . . . . . . Congestion-controlled transport layer protocol . . . . . . . . . . . . . . . . . . . . . Molecular communication and network model . . . . . . . . . . . . . . . . . . . . . An architecture of nanonetwork utilized in [78] . . . . . . . . . . . . . . . . . . . . Envisioned architecture for nano-healthcare . . . . . . . . . . . . . . . . . . . . . . . . Typical Internet of Bio-Nano-Things (IoBNT) used in [89] . . . . . . . . TRouting in heterogeneous communications of BANs . . . . . . . . . . . . . Example DNA packet encoded into the carrier plasmid . . . . . . . . . . . . Proposed model of OR-DMC, ACK, timing sequence, and different types of messages are shown . . . . . . . . . . . . . . . . . . . . . . . . . . Structure of molecular frames and techniques for error in bacterial quorum communications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microfluidic frame carried by continuous phase in a microfluidic channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A general structure of a MaaF frame with n bit slots . . . . . . . . . . . . . . . Communication channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Molecular communication components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Signal molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nanomachine architecture on the cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stochastic model of molecular communication . . . . . . . . . . . . . . . . . . . . . Molecular communication system model. . . . . . . . . . . . . . . . . . . . . . . . . . . . (a) Passive transport and (b) active transport of molecules in the molecular communication environment . . . . . . . . . . . . . . . . . . . . . .

2 3 15 17 25 31 32 34 36 38 40 44 52 53 54 70 71 72 74 75 77 78

xi

xii

List of Figures

Fig. 3.8

Fig. 3.9 Fig. 3.10

Fig. 3.11

Fig. 3.12

Fig. Fig. Fig. Fig. Fig. Fig.

3.13 3.14 3.15 3.16 3.17 3.18

Fig. 3.19

Fig. 3.20 Fig. 3.21

Fig. 3.22

Fig. 3.23

Fig. 3.24

The probability density function of the latency in a semi-infinite interval .−∞, d for various sender and receiver bio-nanomachine distances 2 .d = 1, 2, 4, 8(μm), D = 0, 1(μm /s) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The probability mass function of the latency in a finite interval .[0, d] for various .d = {1, 5, 9} .(μm), D = 0, 1(μm2 /s) . . The probability density function of the latency in a fluid medium for various fluid velocity 2 .v = {0, 0.1, 0.2, 0.4} (μm/s), D = 0, 1(μm /s), and .d = 4(μm) The probability mass function of the latency for various inter-repeater distances .l = {d/2, d/4, d/8} at 2 .N = 2, d = 10(μm) and .D = 0, 1(μm /s). . . . . . . . . . . . . . . . . . . . . . . . . The probability mass function of the latency for various numbers of molecules that a repeater releases 2 .N = {2, 5, 10} at l = d/4, d = 10(μm), and D = 0.1(μm /s) . . Classification of error control: types of noises and techniques . . . . . Representation of a free diffusion-based MC system and noises . . . Representing a cellular signaling-based MC system and noises. . . . Error prevention technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Capacity and path loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Illustration of the MCvD environment where one point T x communicates with one spherical Rx in a three-dimensional environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The number of observed molecules within the transparent Rx at time t versus time t, where 5 2 .Ntx = 10 , rR = 0.5 μm, d = 4 μm, D = 1000 (μm /s) . . . . . . . . Estimation in a cylindrical diffusive MC environment by using ring-shaped Rxs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The MSE of different CIR estimation schemes versus pilot sequence length: LSSE, least squares; ML, maximum likelihood; P-CRLB, pilot-based CRLB; DD-ML, decision-directed ML; DD-LS, decision-directed LS; SB-CRLB, semi-blind CRLB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Molecular communication system and its three core components: the transmitter (T x), the molecular communication channel, and the receiver (Rx) . . . . . . . . . . . . . . . . . . . . . Transmission strategy of 4-MCPM. The first two bits determine the emission time of the molecular signal, while the third and last bit determines the emission intensity of the signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Schematic diagram of a source-sink pair, where different colors represent transmitter-receiver links using different types of molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

82 83

84

85

86 87 88 89 94 94

96

98 102

106

109

116

118

List of Figures

Fig. 3.25 Symbol error probability versus distance between point transmitter and spherical receiver (r − rj ) in a system with single link [136] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 3.26 Probability of symbol error versus threshold T1 for the detection of symbols 0 and 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 3.27 System model of MIMO-MC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 3.28 Transmitter diagram of SM-MC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 3.29 Information can be encoded in the concentration or number of particles released, in their type or structure, or at the time of release . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 3.30 The new communication system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 3.31 Time mechanism of releasing different ions when transmitting bit information “1” or “0” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 3.32 Communication scenarios for nT x = 8 scenario. (a) MISO scenario in 3D view. (b) MISO scenario in 2D view. Black lines show the region boundaries on Rx, and each region-transmitter conjugate is indexed from 0 to 7, consecutively . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 3.33 Natural coding (NC) and gray coding (GC) bit error rate for maximum count decoder (MCD), machine learning (ML), symbol-by-symbol maximum likelihood estimation (MLE) approaches with tb = 0.166s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 3.34 Natural coding (NC) and gray coding (GC) bit error rate for maximum count decoder (MCD), machine learning (ML), symbol-by-symbol maximum likelihood estimation (MLE) approaches with tb = 0.555s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 3.35 The system model for 8 × 8 diffusion-based molecular MIMO system. The distance between the nearest points of T x and Rx antennas is denoted by dtr , and the distance between the projection of the spherical receiver antenna and the center of UCA is represented by dru . The radius of the spherical absorbing Rx antenna is given by rr . . . . . . . . . . . . . . . Fig. 3.36 BER versus MT plots for the underlying 8 × 8 DB-MoMIMO system for the existing and proposed 2 modulation schemes; D = 79.4 μ ms , dtr = 4 μm, dru = 6.2 μm, rr = 4 μm, Tb = 0.6, input bit stream length = 60, and L = 6 [146] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 3.37 BER versus Tb plots for the underlying 8 × 8 DB-MoMIMO system for the existing and proposed 2 modulation schemes; D = 79.4 μ ms , dtr = 4 μm, dru = 6.2 μm, rr = 4 μm, MT = 100, input bit stream length = 60, and L = 6 [146] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 3.38 Signaling sublayer in molecular communication based on Shannon’s model of communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xiii

122 122 123 124

132 139 142

143

144

145

147

148

149 154

xiv

Fig. 4.1 Fig. 4.2 Fig. 4.3

List of Figures

NanoNS3 architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Generic IoBNT architecture for ubiquitous healthcare . . . . . . . . . . . . . Gene delivery process described using a four-layer communication protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 4.4 HIV infection-layered protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 4.5 IoT-ATN-layered architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 4.6 DNA- and enzyme-based protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 4.7 Molecular layered architecture for illustrating the flow of molecules information from the T x to Dx by using relay .(Rx) . . . Fig. 4.8 Example OSI to nanoscale communication network mapping . . . . . Fig. 4.9 Layered network communication model of the communication of information from the DTE to the GA . . . . . . . . . . . Fig. 4.10 Representation of the MI in relation to the velocity of drift and the diffusion coefficient [53] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 4.11 Internetwork equivalence of gene expression . . . . . . . . . . . . . . . . . . . . . . .

168 170 172 174 176 178 179 181 189 191 192

List of Tables

Table 2.1 Main functionalities and tasks provided by each layer in molecular communications following the model described by [7] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Table 3.1 Concentration-based modulation techniques for molecular communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Table 3.2 Type-based techniques for molecular communication . . . . . . . . . . . . . . Table 3.3 Timing-based techniques for molecular communication . . . . . . . . . . . Table 3.4 Slot durations and average emitted molecules per symbol for single-type molecule mcvd schemes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Table 3.5 Ideal data rate for different modulation schemes in MC . . . . . . . . . . . Table 3.6 Binary bit addition modulation technique depicting molecule transmission scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Table 3.7 Mapping strategy for selecting current antenna index based on previously activated antenna or previous state (P.S.) and current grouped input bits (i/p) in 8 × 8 MoMIMO system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Table 4.1 IEEE 1906.1 core component specification for 1906.1.1 standard for nanoscale communication systems . . . . . . . . . . . . . . . . . . . .

16 111 113 115 117 126 130

136 183

xv

Chapter 1

Introduction

Molecular communication (MC) is described by Akyildiz et al. [1] as an important field with prospective applications, particularly in human health. As a result, [1] demonstrate how MC theory may facilitate the perception and regulation of this information as it spreads in vivo through the biochemistry supporting the patient’s body, its cells, and their molecular makeup. For example, it is thought that errors in the transmission of the molecular information that controls cell differentiation and growth may play a role in the start and spread of cancer. Additionally, two fundamental routes for applying the MC theory to human health, those of natural or artificially created systems, are identified in [1], as further shown in Fig. 1.1 [2–9]. The MC theory is used to model how information flows at the following scales in the top-down paradigm (natural system direction) shown in Fig. 1.1: (1) the body system scale, where organs and tissues are connected to perform a specific function; (2) the cellular scale, where cells work together to process information; and (3) the molecular scale, where information is stored in substances and exchanged through chemical reactions and molecule transport. In the bottom-up paradigm (synthetically engineered systems) depicted in Fig. 1.1, the MC theory is used to design the following: (1) MCs components using genetic programming tools from synthetic biology and technologies to interface with classical electronics; (2) MCs devices, which are engineered biological systems, from microbes to human cells, capable of preprogrammed MC behaviors; and (3) MC networks, where engineered biological systems connect to get the desired MC behaviors [2]. As investigated since their beginnings (about 15 years ago), MCs essentially analyze the transmission of information at the nanolevel; the cells, the “smart” devices that constitute our bodies, principally communicate fundamentally through the transport and binding of molecules. Nature has contrived a way for all these transmission of biomolecules to act in concert so efficiently and consistently, despite their noisy environment and principally diffusive transport. This is an astounding achievement in robust communication. Scientists need to figure out how it works so they can make devices that can talk to nature in its own language [10–16]. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 Y. Cevallos et al., Molecular Communications, https://doi.org/10.1007/978-3-031-36882-0_1

1

2

1 Introduction

Fig. 1.1 Two main directions for applying molecular communication theory to human health

Unlike current telecommunication paradigm, in MCs, molecules are used as information carriers: sender biological nanomachines, also known as bionanomachines, encode information on molecules and release them into the environment. The molecules then spread to receiver biological nanomachines through the environment, where the latter decode the information by interacting

1 Introduction

3

biochemically with the molecules. Sender-receiver systems abound in biology, with communication systems sending information in various forms [17]. Information and communication technology (ICT) provides a quantitative basis for analyzing these processes and is being applied to study natural and synthetic genetics. Senderreceiver systems with cells and biomolecules can be seen as the information processing agents [18–21]. This processing allows scientists to quantify signaling, and how this is the first glimpse into Shannon’s predictions in biological systems (Fig. 1.2). The upper panel of Fig. 1.2 visualizes communication layers as a multilevel jigsaw puzzle, working between factors such as proteins, cells, and organs. Shannon’s information theory can be used to quantify the information flow in all such sender-receiver systems [10, 11, 22–24]. In the lower panel of Fig. 1.2, inputs from the sender S vary in terms of dose (e.g., chemical concentration), kind of biomolecule (e.g., AHL (acyl-homoserine lactone), volatile aldehyde, dopamine, or even DNA (deoxyribonucleic acid) fragments), and rate of creation. Information is transferred across the channel. The bandwidth B is the range of frequency permitted by the channel (the change in molecule concentration; Hz); the channel capacity C is modulated by the equation given

Fig. 1.2 Information theory paradigm to analyze molecular communications

4

1 Introduction

and measured in bits per second; and the signal and noise components are S and N, respectively. The receiver R uses appropriate receivers, such as cell surface receptors. The “meaning” of the signal is extracted by a modulation system, such as a cell signaling route, which connects the signal to the interpreter (e.g., a responsive promoter for gene expression). The responses of the outputs, such as gene expression, are measured relative to space, time, and input dosage [22]. Then, as it seems inertial to note, MCs can be described as the use of molecules as messages between the transmitter and the receiver. This mode of communication is the most promising for general applications. An example of this communication method is the communication between neighboring cells in the human body, which is conducted using the diffusion of different types of molecules that encode different types of messages. Molecular communication allows biological and artificially created nanomachines to communicate using molecules over a short or long nanodistance. In MC, senders encode information onto molecules (called information molecules). Information molecules are then loaded onto carrier molecules and propagate to a receiver. Upon receiving the information molecules, the receiver reacts biochemically to the incoming information molecules [25, 26]. Carrier molecules, neurotransmitters, hormones, molecular motors, viruses, and more are being discovered and studied. The information molecules are proteins, ions, or even DNA strands. In MCs, there are several different ways that molecules move from one network node to another through the communication channel. Because this mode of communication is inspired by natural phenomena, some current research topics include calcium signaling, chemical molecule diffusion, molecular motors, bacteria signaling, flagellated bacteria communication, and pheromone signaling [25, 27–34]. At this point, the reader may be curious as to why we direct the analysis of MCs specifically toward human health. The answer to this question is explained in detail on the last page of this chapter. It has to do with a terrible and unexpected event that is hurting people right now. Each year, hundreds of millions of people are affected by viral infections, but many do not have vaccines or effective treatment during or after infection. The COVID-19 pandemic has brought attention to this problem. It showed how quickly viruses can spread and how they can affect society as a whole. To fight viral infections and possible future pandemics, we need to come up with new methods that combine different fields of knowledge. In the past decade, an interdisciplinary area involving bioengineering, nanotechnology, and ICT has been developing, known as MCs [11, 23, 35–45]. This new emerging area uses elements of classical communication systems (as we have described in the previous paragraphs) and maps them to molecular signaling and communication found inside and outside the body, where the aim is to develop new tools that can serve future medicine. The information and communication technology framework provides a novel perspective to fight human diseases [37, 46–49]. One of these tools is the ability to characterize the signaling processes between cells and infectious disease locations at various levels of the human body [50]. As a result, MCs have emerged as a link between communication engineering and networking, as well as molecular biology and bioengineering [51–53]. Infection

1 Introduction

5

theranostics, the term for the development of novel technologies that can result in effective treatment or diagnosis of infectious diseases, is mostly at the junction of subjects like bioengineering, material sciences, and medical sciences (therapydiagnostics). By incorporating communications features of transmission, propagation, and reception, MCs can be crucial in these new technologies. For example, organ-on-a-chip systems based on microfluidics offer experimental models of the transmission, propagation, and receipt of molecules from cell-cell, tissue-tissue, and even organ-organ, with or without other external molecular agents. Another illustration is the detection of airborne viruses by biosensors that can be either integrated into all-in-one devices or connected with suggested new 6G (sixthgeneration wireless) infrastructures, such as the intelligent reflecting surface. In this situation, the development of such infrastructures via electrically charged droplets or binding-ligand proteins can entail modeling expertise from MCs researchers [51]. Information technology may advance by incorporating bio-nanomachines into currently accessible silicon-based electrical systems employing MCs. A bionanomachine is a nano-to-micro-scale device consisting of biological components, capable of simple chemical functions [54]. Future cell phones might, for instance, come equipped with bio-nanomachines that use MC to instantly evaluate biochemical signals (such as blood or sweat molecules) on the chip. Another method MCs could advance information technology is through a dermal display panel. A dermal display screen would consist of three billion bio-nanomachines that are implanted under a person’s skin and communicate with one another. The Internetof-nano-things, which would consist of widely dispersed bio-nanomachines with MCs, might also be created, as could body-area nanonetworks [54, 55]. Bioengineering is focused on developing revolutionary new technologies for a civilization that bears the promise of subtle sensing and actuation capabilities inside the human body through a network of micro- and nano-sized devices in this fantastical and futuristic paradigm. In order to connect and communicate with the human body, these devices can make use of the natural signals already present in cells and tissues. The primary benefit is the potential to improve implantable systems’ biocompatibility through the incorporation of synthetic biology. For both acquiring new insights into the viral traits and characteristics and for developing innovative treatments, this novel study area has the potential to be crucial in preventing both present and future pandemics. MCs can therefore primarily contribute to the following areas [51]: 1. To the characterization of the virus propagation within the body. 2. To understand the mechanism used by the virus to enter the human body or the mechanism of expulsion. 3. To understand how the airborne virus propagates in the air. Consequently, infection by a virus can be analyzed through conventional communication systems models (i.e., from Shannon’s perspective, with communication systems composed of a transmitter, a communication channel, and a receiver). Particularly, the models to analyze a virus depend on in-body and out-body circumstances.

6

1 Introduction

In-Body MCs Models The MCs paradigm explains in detail how the virus behaves and spreads over time inside the body. The virions are viewed as information carriers in communications because they carry messages (genome) from the transmission place to the reception location, which can be host cells in particular organs or tissues. The activity of infection is the information that the virions transmit [51]. Out-Bodied MCs Models The primary method of virus transmission between people is by airborne infection. Viruses spread to another person by breathing themselves into their lungs once they have been released into the air. This process makes it possible for the virus infection to quickly advance to local or pandemic proportions. Other methods of virus transmission between people (such as sexual or human-to-human contact) exist [51]). Then, fresh insights from tried-and-true MC methods can be applied to infectology and assessed from the perspective of healthcare. Additionally, technologies developed in communication and information theory can be used to perform a reverse check on proven infection prevention methods [56]. These evaluations will demonstrate whether the parallels and dualities can result in novel preventative strategies that are effective in the actual world and lower infection rates. Due to the maturity of both research fields, it is now possible to examine each other’s respective other sides [52]. With this dual perspective in [52] is studied coronavirus infectious disease transmission via aerosol propagation as viewed from a communication and information theory perspective. Work is done to identify the dualities and parallels between the transmission of infectious particles and macroscopic air-based MC. Users are depicted as mobile nodes in the notion for the first time in a multiuser setting. Healthy users may contract an infection with a certain likelihood if they are exposed to enough viral load by viral-sensitive aperture areas, such as spatially distributed antennas in the wireless communications industry. Infected users create aerosol clouds in the sense of broadcasting. These users may later develop into pathogenladen spreaders or superspreaders in relation to store and forward relaying after a period of incubation. Similar to reliability information in MCs, the viral load frequently affects the course of the disease in addition to being a pure binary dilemma of contracting or avoiding infection. The objective is to reduce the reciprocal information between infected and noninfected users, where the information corresponds to infected and non-infected particles, in the sense of Shannon’s information theory. The likelihood that a particular density of pathogen-laden particles effectively make contact with the receiver’s effective aperture area is directly correlated with the likelihood of infection. After a specific incubation period, the receiver will get infected with a specific probability if the receiver-side density exceeds the user-dependent infection threshold [11, 52, 57]. This situation can be modeled mathematically using mutual information. Start with a point-to-point channel. Let x be the symbol for the channel’s input, and let y be the corresponding observation. The mutual information is a function of the channel input distribution, .p(x), and the joint probability, .p(x; y). The distribution

1 Introduction

7

of channel inputs depends on the breathing events, and the distribution of channel inputs and outputs together depends on how turbulent the atmospheric channel is. This concept can be generalized to multiuser channel models. Three crucial aspects are illustrated by the comparison between mutual information and aerosolbased infection: the risk of infection depends on the input channel distribution, the turbulence channel, and the infection threshold [52]. Numerous actions relating to channel input distribution, channel propagation, and channel output properties are present in this stochastic method to infection control. For instance, protective masks have an impact on the distribution of channel input. Channel propagation is primarily affected by spatial and temporal actions. The infection threshold at the channel output is influenced by biochemical processes and physiological states. The several users’ spatially dispersed, temporally variable aerosol clouds are subject to a dynamic channel with turbulences and shifting characteristics. Different impacts alter the channel, causing the aerosols to move through space at varying speeds and distances. The user’s movement, ventilation systems, airstreams introduced by windows and doors, temperature gradients, and meteorological occurrences are the main causes of air turbulence. By positively altering the propagation vector, turbulence can either extend aerosols’ range or shorten it as it pushes aerosols toward the earth. By adding an external fan to the arrangement to create an artificial air movement, turbulence in the testbed environment can be modeled. Shannon’s model helps explain the turbulent channel [52]. Also, in [36, 54], the authors looked at the mechanisms of human airborne pathogen transmission and reception for indoor environments from the standpoint of MCs. The use and adaptation of well-known communication engineering techniques to model the transmission of infectious diseases between humans result from this MCs perspective. A proposed end-to-end system model takes into account the pathogen-filled cough and sneeze droplets as the input and the human infection state as the output. This model proposes a receiver model that employs the middle of the human face as the reception interface and leverages gravity, starting velocity, and buoyancy to propagate droplets. Additionally, the quantity of spreading droplets is modeled as a random process to calculate the chance of infection for a human who is not already affected. The numerical outcomes show that the human’s sex and the duration of exposure have an impact on the likelihood of infection. Additionally, the safe coughing angle for a human to infect fewer individuals should be less than .−25 degrees, and the social distance for a horizontal cough should be at least 1.7 meters. The fundamental legacy is that of Shannon, where a purely syntactic characterization of information ranks systems based on their maximum information efficiency. Shannon’s work has the potential to provide new insights into life processes and ways to interact with and manage them. The latter measures appear inappropriate for biological systems, since differing chances of survival can emerge from unequal information transmission and storage (bearing distinct meanings), based on a mathematical abstraction capable of capturing the characteristics and actions of a population of single-celled organisms whose survival is tied to information retrieval from the environment [58].

8

1 Introduction

Shannon’s mathematical theory of communication’s conceptual success and applicability within the developing fields of electronics and telecommunications depended on its restriction to strictly syntactic considerations, where the meaning of information is considered irrelevant to the engineering problem. The study of mathematical algorithms where information can be represented by numbers has been largely based on the idea of information entropy as a universal metric to quantify information, or its absence, i.e., noise, and the definition of a communication channel’s capacity as the theoretical maximum amount of information, or mutual information (MI), which can be unambiguously propagated between two points in space (transmission), or time (storage). The syntactic character of information theory presents a barrier to its application to living systems, despite attempts to do so in biology, from neurology to biochemistry, and data analytics for bioinformatics, even abstracting biological systems as communication channels. It makes sense that some communications in biological systems be “more significant” than others. Although attempts have been made to give this concept a quantitative foundation, there is still no overall structure [58]. As a result, the authors in [58] create a model, which they refer to as a computational state machine. This model is then applied in a simulation framework designed specifically to reveal the emergence of a “subjective information,” or a trade-off between a living system’s ability to maximize the acquisition of information from the environment and the maximization of its growth and survival over time. As observed, MCs analysis from ICT (with generic or own communications models) [10, 59] could enhance the human life quality due to the enormous advantages that mean the knowledge of vital mechanisms that have facilitated the evolution of organisms through millions of years. These enhancements could be applied to various areas such as environment, military, materials science, industry, computers, and communication sciences, and for sure the medical field [60–62]. In this last field, MC theories could avoid diseases or improve the treatment of diseases [63–68]. It can be used for military NBC (nuclear, biological, and chemical) surveillance, and defense. When combined with chemical processes, MC systems can create novel molecular configurations and structures in the industrial setting. Advanced nanomaterials can also be included into nanomachines to create antibacterial, antifouling, and mosquito-repellent textiles when they communicate with one another via MC. It can be used to recognize and keep an eye on a few certain molecules that can lead to environmental problems including illegal pollution and radioactive leaks. It can be employed in biomedicine for a variety of purposes, including biological hybrid transplantation, medication delivery systems, illness therapy, and health supervision [69, 70]. The explained and potential applications referred to in earlier paragraphs are the reason (and the answer to our previous question) to currently have a permanent research effort in MCs. Thus, in this investigation, we utilize the conventional theories of digital communication systems (specifically the analysis of the layers’ tasks in digital networking) in MCs systems.

References

9

References 1. I.F. Akyildiz, M. Pierobon, S. Balasubramaniam, Moving forward with molecular communication: From theory to human health applications [point of view]. Proc. IEEE 107(5), 858–865 (2019) 2. A. Zadeh Shirazi, E. Fornaciari, M.D. McDonnell, M. Yaghoobi, Y. Cevallos, L. TelloOquendo, D. Inca, G.A. Gomez, The application of deep convolutional neural networks to brain cancer images: a survey. J. Pers. Med. 10(4), 224 (2020) 3. D. Romanchenko, M. Tartie, B. Q. Le, J. T. Gómez, F. Dressler, Molecular communication channel modelling in FPGA technology. Electron. Commun. EASST 80, (2021). http://doi. org/10.14279/tuj.eceasst.80.1144 4. I. Isik, M. B. Er, E. Isik, Analysis and classification of the mobile molecular communication systems with deep learning. J. Ambient Intell. Humaniz. Comput. 13(5), 2903–2919 (2022) 5. Z. Cheng, J. Sun, J. Yan, Y. Tu, Optimizations for mobile mimo relay molecular communication via diffusion with network coding. KSII Trans. Internet Inf. Syst. (TIIS) 16(4), 1373–1391 (2022) 6. I. Isik, How mobility of transmitter and receiver affects the communication quality. AIP Adv. 12(2), 025205 (2022) 7. Y. Bayraktar, E. Isik, I. Isik, A. Ozyilmaz, M. Toprak, F. Kahraman Guloglu, S. Aydin, Analyzing of alzheimer’s disease based on biomedical and socio-economic approach using molecular communication, artificial neural network, and random forest models. Sustainability 14(13), 7901 (2022) 8. C. Fonseca, M. T. Barros, A. Odysseos, S. Kidambi, S. Balasubramaniam, Quasi-Spherical Absorbing-Reactive Receiver Model of Glioblastoma Cells for Exosome-based Molecular Communications 9. D. Vimalajeewa, S. Balasubramaniam, D. P. Berry, G. Barry, Virus particle propagation and infectivity along the respiratory tract and a case study for SARS-CoV-2. Sci. Rep. 12(1), 1–14 (2022) 10. K. R. Pilkiewicz, P. Rana, M. L. Mayo, P. Ghosh, Molecular communication and cellular signaling from an information-theory perspective, in Nanoscale Networking and Communications Handbook (2019), pp. 235–256 11. D. Bi, A. Almpanis, A. Noel, Y. Deng, R. Schober, A survey of molecular communication in cell biology: Establishing a new hierarchy for interdisciplinary applications. IEEE Commun. Surv. Tutorials 23(3), 1494–1545 (2021) 12. N. Dalchau, G. Szép, R. Hernansaiz-Ballesteros, C. P. Barnes, L. Cardelli, A. Phillips, A. Csikász-Nagy, Computing with biological switches and clocks. Nat. Comput. 17(4), 761– 779 (2018) 13. L. Khaloopour, M. Mirmohseni, M. Nasiri-Kenari, Joint sensing, communication and localization of a silent abnormality using molecular diffusion. Preprint (2022). arXiv:2203.16641 14. S. Figueiredo, N. Souto, F. Cercas, Low-complexity channel codes for reliable molecular communication via diffusion. Sensors 22(1), 41 (2021) 15. F. Vakilipoor, F. Ratti, H. Awan, M. Magarini, Low complexity receiver design for time-varying Poisson molecular communication channels with memory. Digit. Signal Process. 124, 103187 (2022) 16. L. Chouhan, M.-S. Alouini, Rescaled brownian motion of molecules and devices in threedimensional multi-user mobile molecular communication systems. IEEE Trans. Wireless Commun. 21(12), 10472–10488 (2022) 17. M. Kuscu, O.B. Akan, Detection in molecular communications with ligand receptors under molecular interference. Digit. Signal Process. 124, 103186 (2022) 18. J. T. Gómez, K. Pitke, L. Stratmann, F. Dressler, Age of information in molecular communication channels. Digit. Signal Process. 124, 103108 (2022) 19. S. Sarkar, S. Choubey, Efficacy of information transmission in cellular communication. Preprint (2020). arXiv:2010.05291

10

1 Introduction

20. S.S. Somathilaka, D. P. Martins, S. Balasubramaniam, Information flow of cascading bacterial molecular communication systems with cooperative amplification, in IEE ICC 2022 (IEEE, 2022) 21. Z. Cheng, J. Yan, J. Sun, Y. Tu, K. Chi, Joint optimizations of relays locations and decision threshold for multi-hop diffusive mobile molecular communication with drift. IEEE Trans. NanoBiosci. 21(3), 454–465 (2022) 22. T. Nakano, M. Moore, A. Enomoto, T. Suda, Molecular communication technology as a biological ICT, in Biological Functions for Information and Communication Technologies (Springer, 2011), pp. 49–86 23. M.J. Salariseddigh, U. Pereg, H. Boche, C. Deppe, V. Jamali, R. Schober, Deterministic identification for molecular communications over the Poisson channel. Preprint (2022). arXiv:2203.02784 24. V. Jamali, H. M. Loos, A. Buettner, R. Schober, H. V. Poor, Olfaction-inspired mcs: Molecule mixture shift keying and cross-reactive receptor arrays. Preprint (2022). arXiv:2203.04225 25. D. C. Ferreira, L. P. Reis, N. V. Lopes, A nanocommunication system for endocrine diseases. Clust. Comput. 20(1), 689–706 (2017) 26. S. Ahmed, J. Hu, S. M. Naqvi, Y. Zhang, L. Linze, A. M. Iderawumi, Molecular communication network and its applications in crop sciences. Planta 255(6), 1–12 (2022) 27. I. F. Akyildiz, F. Fekri, R. Sivakumar, C. R. Forest, B. K. Hammer, Monaco: fundamentals of molecular nano-communication networks. IEEE Wireless Commun. 19(5), 12–18 (2012) 28. M. Femminella, G. Reali, Implementation issues of diffusion-based molecular communications receivers based on transcriptional elements. Digit. Signal Process. 124, 103160 (2022) 29. M. Egan, B. C. Akdeniz, B. Q. Tang, Stochastic reaction and diffusion systems in molecular communications: Recent results and open problems. Digit. Signal Process. 124, 103117 (2022) 30. L. Chouhan, P. K. Upadhyay, P. K. Sharma, A. M. Salhab, On anomalous diffusion of devices in molecular communication network. Preprint (2022). arXiv:2203.14808 31. M. Rezaei, H. K. Rudsari, M. Javan, N. Mokari, E. A. Jorswieck, M. Orooji, Molecular communication transmitter design in limited-capacity storage regime. IEEE Trans. NanoBiosci. 22(2), 212–222 (2022) 32. Z. Ye, O. Elishco, Codes over absorption channels. Preprint (2023). arXiv:2302.09842 33. H. Arjmandi, M. Zoofaghari, M. Rezaei, K. Kanebratt, L. Vilen, D. Janzen, P. Gennemark, A. Noel, Diffusive molecular communication with a spheroidal receiver for organ-on-chip systems. Preprint (2023). arXiv:2302.09265 34. S. Lotter, M. T. Barros, R. Schober, M. Schäfer, Signal reception with generic threestate receptors in synaptic MC, in GLOBECOM 2022-2022 IEEE Global Communications Conference (IEEE, 2022), pp. 4529–4534 35. A. Gohari, M. Mirmohseni, M. Nasiri-Kenari, Information theory of molecular communication: Directions and challenges. IEEE Trans. Mol. Biol. Multi-Scale Commun. 2(2), 120–142 (2016) 36. T. Khan, M. Civas, O. Cetinkaya, N. A. Abbasi, O. B. Akan, Nanosensor networks for smart health care, in Nanosensors for Smart Cities (Elsevier, Amsterdam, 2020), pp. 387–403 37. X. Chen, M. Wen, F. Ji, Y. Huang, Y. Tang, A. W. Eckford, Detection interval for diffusion molecular communication: How long is enough? Preprint (2022). arXiv:2204.08636 38. I. Iqbal, M. Nazir, A. Sabah, Design of energy-efficient protocol stack for nanocommunication using greedy algorithms. J. Comput. Networks Commun. 2022, Article ID 3150865 (2022) 39. D. Jing, Y. Li, A. W. Eckford, An extended kalman filter for distance estimation and power control in mobile molecular communication. Preprint (2022). arXiv:2205.09596 40. A. Awad, M. M. Fouda, M. M. Khashaba, E. R. Mohamed, K. M. Hosny, Utilization of mobile edge computing on the internet of medical things: A survey. ICT Exp. 9(3), 473–485 (2022) 41. O. Yetimoglu, M. K. Avci, B. C. Akdeniz, H. B. Yilmaz, A. E. Pusane, T. Tugcu, Multiple transmitter localization via single receiver in 3-d molecular communication via diffusion. Digit. Signal Process. 124, 103185 (2022) 42. Y. Li, L. Lin, W. Guo, D. Zhang, K. Yang, Error performance and mutual information for IoNT interface system. IEEE Internet Things J. (2022)

References

11

43. W. Pan, X. Chen, X. Yang, N. Zhao, L. Meng, F. H. Shah, A molecular communication platform based on body area nanonetwork. Nanomaterials 12(4), 722 (2022) 44. M. Egan, M. Kuscu, M. T. Barros, M. Booth, A. Llopis-Lorente, M. Magarini, D. P. Martins, M. Schäfer, P. Stano, Toward interdisciplinary synergies in molecular communications: Perspectives from synthetic biology, nanotechnology, communications engineering and philosophy of science. Life 13(1), 208 (2023) 45. J. Yoon, J. Lim, M. Shin, T. Lee, J.-W. Choi, Toward bioelectronic device based on bionanohybrid composed of nanomaterials and biomaterials: From nucleic acid and protein to living cell. Appl. Phys. Rev. 10(1), 011302 (2023) 46. C. Koca, M. Civas, S. M. Sahin, O. Ergonul, O. B. Akan, Molecular communication theoretical modeling and analysis of SARS-CoV2 transmission in human respiratory system. Preprint (2020). arXiv:2011.05154 47. T. Nakano, M. J. Moore, F. Wei, A. V. Vasilakos, J. Shuai, Molecular communication and networking: Opportunities and challenges. IEEE Trans. Nanobiosci. 11(2), 135–148 (2012) 48. L. F. Borges, M. T. Barros, M. Nogueira, Toward reliable intra-body molecular communication: An error control perspective. IEEE Commun. Mag. 59(5), 114–120 (2021) 49. Y. Sun, H. Bian, Y. Chen, A photolysis-assist molecular communication for tumor biosensing. Sensors 22(7), 2495 (2022) 50. J. Wang, M. Peng, Y. Liu, X. Liu, M. Daneshmand, Performance analysis of signal detection for amplify-and-forward relay in diffusion-based molecular communication systems. IEEE Internet Things J. 7(2), 1401–1412 (2019) 51. M. T. Barros, M. Veleti´c, M. Kanada, M. Pierobon, S. Vainio, I. Balasingham, S. Balasubramaniam, Molecular communications in viral infections research: Modelling, experimental data and future directions. Preprint (2020). arXiv:2011.00002 52. M. Schurwanz, P. A. Hoeher, S. Bhattacharjee, M. Damrath, L. Stratmann, F. Dressler, Infectious disease transmission via aerosol propagation from a molecular communication perspective: Shannon meets coronavirus. Preprint (2020). arXiv:2011.00290 53. T. Nakano, Molecular communication: A 10 year retrospective. IEEE Trans. Mol. Biol. MultiScale Commun. 3(2), 71–78 (2017) 54. T. Nakano, T. Suda, Y. Okaie, M. J. Moore, A. V. Vasilakos, Molecular communication among biological nanomachines: A layered architecture and research issues. IEEE Trans. Nanobiosci. 13(3), 169–197 (2014) 55. Z. Sakkaff, B. D. Unluturk, M. Pierobon, Applying molecular communication theory to multiscale integrated models of biological pathways, in Proceedings of the Sixth Annual ACM International Conference on Nanoscale Computing and Communication (2019), pp. 1–2 56. J.-T. Huang, H.-Y. Lai, Y.-C. Lee, C.-H. Lee, P.-C. Yeh, Distance estimation in concentrationbased molecular communications, in 2013 IEEE Global Communications Conference (GLOBECOM) (IEEE, 2013), pp. 2587–2591 57. P. A. Hoeher, M. Damrath, S. Bhattacharjee, M. Schurwanz, On mutual information analysis of infectious disease transmission via particle propagation. Preprint (2021). arXiv:2101.12121 58. T. S. Barker, M. Pierobon, P. J. Thomas, Subjective information and survival in a simulated biological system. Entropy 24(5), 639 (2022). [Online]. Available: https://www.mdpi.com/ 1099-4300/24/5/639 59. Y. Cevallos, L. Tello-Oquendo, D. Inca, N. Samaniego, I. Santillán, A. Z. Shirazi, G. A. Gomez, On the efficient digital code representation in dna-based data storage, in Proceedings of the 7th ACM International Conference on Nanoscale Computing and Communication, ser. NanoCom ’20 (Association for Computing Machinery, New York, NY, USA, 2020). [Online]. Available: https://doi.org/10.1145/3411295.3411314 60. A. K. Shrivastava, D. Das, R. Mahapatra, S. P. Mohanty, dMole: A novel transreceiver for mobile molecular communication using robust differential detection techniques. IEEE Trans. NanoBiosci. 19(4), 609–621 (2020) 61. R. Paridar, N. Mokari, E. Jorswieck, M. R. Javan, On the reception process of molecular communication-based drug delivery. Preprint (2021). arXiv:2112.12485

12

1 Introduction

62. S. Senturk, I. Kok, F. Senturk, Internet of nano, bio-nano, biodegradable and ingestible things: A survey. Preprint (2022). arXiv:2202.12409 63. Y. Cevallos, L. Molina, A. Santillán, F. De Rango, A. Rushdi, J. B. Alonso, A digital communication analysis of gene expression of proteins in biological systems: A layered network model view. Cogn. Comput. 9(1), 43–67 (2017) 64. Y. Cevallos, L. Tello-Oquendo, D. Inca, C. Palacios, L. Rentería, Genetic expression in biological systems: A digital communication perspective. Open Bioinf. J. 12(1), 45–49 (2019) 65. Y. Cevallos, L. Tello-Oquendo, D. Inca, D. Ghose, A. Z. Shirazi, G. A. Gomez, Health applications based on molecular communications: A brief review, in 2019 IEEE International Conference on E-health Networking, Application & Services (HealthCom) (IEEE, 2019), pp. 1–6 66. Y. Cevallos, T. Nakano, L. Tello-Oquendo, A. Rushdi, D. Inca, I. Santillán, A. Z. Shirazi, N. Samaniego, A brief review on DNA storage, compression, and digitalization. Nano Commun. Networks 31, 100391 (2022) 67. Y. Cevallos, T. Nakano, L. Tello-Oquendo, D. Inca, I. Santillán, A. Z. Shirazi, A. Rushdi, N. Samaniego, Modeling gene expression and protein delivery as an end-to-end digital communication system. Open Bioinf. J. 14(1), 21–35 (2021) 68. Y. Cevallos, T. Nakano, L. Tello-Oquendo, N. Chopra, A. Z. Shirazi, D. Inca, I. Santillán, Theoretical basis for gene expression modeling based on the IEEE 1906.1 standard, in International Conference on Bio-inspired Information and Communication Technologies (Springer, 2021), pp. 145–162 69. X. Wang, Z. Jia, A new modulation method for diffusion molecular communication. Acad. J. Eng. Technol. Sci. 4(2), 7–12 (2021) 70. H. K. Rudsari, M. Zoofaghari, M. Veletic, J. Bergsland, I. Balasingham, The end-to-end molecular communication model of extracellular vesicle-based drug delivery. Preprint (2022). arXiv:2207.01875

Chapter 2

Analysis of Layer’s Tasks in Molecular Communication: Application, Transport, Network, and Link Layers

The most widely used frameworks for describing communications in digital networks are the open systems interconnection (OSI) model and transmission control protocol/Internet Protocol (TCP/IP), with the first one being viewed from a theoretical paradigm (model) and the second from both a theoretical and practical point of view (architecture). A protocol created for that layer is used for communication between entities at different communication devices. The layers oversee tasks including interacting with applications that require communication, connecting devices, handling transmission errors, and actually sending bits over the communication channel. For the design of molecular communication (MC) systems, these frameworks have already been taken into account [1–17], and in this way, Computer network researchers viewed MC as a new communication paradigm, and MC has become a recognized discipline in computer networks in which the models and architectures are the support for models and tools (e.g., information theory) to MC research [8, 18–21]. However, the use of the mentioned networking paradigm for MCs, i.e., the utilization of a layered architecture, does not match exactly; the inertial reason that explains this situation is related to the fact that was connecting electronic devices in a telecommunication system is not the same that the interconnection of biological entities [22, 23], where each device has different limitations, and requirements [24, 25]. Even in the commented situation, molecular systems’ natural behavior has many communicational similarities with conventional communication systems. Thus, surprisingly some facts in MC systems seem to resemble some kind of routing and addressing, the establishment of links, transport of information, and feedback from the destination to sender, almost precisely as if these MCs components were tiny network devices (a single bio-machine1 has very limited capabilities; the interconnection of bio-nanomachines in a nanonetwork can extend their potentials to realize many sophisticated applications [26]), and hence the actual 1 In

Chap. 3 will be explained the definition of a nanomachine.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 Y. Cevallos et al., Molecular Communications, https://doi.org/10.1007/978-3-031-36882-0_2

13

14

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

base for supporting every research in MCs is the perspective of communication engineering [22, 27–35]. Biological cells have the same ability to transmit, receive, and process information as networked computing equipment. The creation of communication protocols for nano-based networks will inspire groundbreaking future uses of bio nanodevices, just as data communication protocols led to the quick expansion and ubiquity of networked computing devices and applications. These two technologies working together have a huge range of potential uses, especially in the medical field where nanoscale devices can execute operations. A variety of elements found in living cells are essential to networked communication. Its applications to medical and health domains are highly anticipated [12, 36–50]. It is crucial to look at MCs from the perspective of digital networks since, for example, one of the most important functions of the data link layer is to convert an imperfect channel into a line free of transmission errors or notify the upper layer of unresolved issues [15, 51]. This approach applied to MCs could play a critical role in facilitating the health applications (among others) development [52], for example, in targeted drug delivery systems to define the effectiveness in cancer treatments [53–55]. The use of a stack of layers in digital networks applied to MCs is so crucial that IEEE (Institute of Electrical and Electronics Engineers) has worked on two standards (1906.1 [56] and 1906.1.1 [57]) in the nano communications area (including communications at the molecular level); in these standards IEEE defines functions of “molecular layers” in a molecular network stack (we analyze these standards in Chap. 4). With the previous explanation, we analyze layers’ functionalities in molecular networks from the digital networking paradigm in this chapter. Besides, in this chapter, investigations that use these molecular functionalities are mentioned.

2.1 Molecular Communication Network Architecture As we mentioned before, the architecture for MCs may be discussed from the Transmission Control Protocol/Internet Protocol (TCP/IP) reference model and architecture and the International Standards Organization/Open Systems Interconnection (ISO/OSI) reference model. Nevertheless, as in typical digital communication networks, a hybrid model (i.e., application, transport, network, link, and physical layers) is commonly utilized [58]. We also employ this hybrid model for MCs supported by [7]. Thus, Fig. 2.1 illustrates how information in a molecular WAN (wide area network) may flow from a source through a router to the destination, representing groups of sender bio-nanomachines, bio-nanomachines with routing functionality, and receiver bio-nanomachines, respectively [7, 8]. The source and router, as well as the router and destination, are within a communication range (short or long nanorange [15]), meaning that information molecules can propagate from one to the other within a reasonable amount of time to induce intended reactions. Similar

2.1 Molecular Communication Network Architecture

15

Fig. 2.1 The flow of information through a molecular communication network is modeled after the WAN architecture

to the TCP/IP reference model, the application layer provides a set of options to implement applications; the transport, network, and link layers provide mechanisms to transmit information over and within a communication range; and the physical layer provides biophysical mechanisms for transmission, propagation, and reception of information molecules over physical media [8]. In this WAN communication, the router selects a communication channel, ensures that the channel is available, and transmits and propagates a type of information molecule to the destination [9]. A layered architecture, like typical digital networks, breaks down a big system into a number of smaller parts (i.e., layers) that are functionally independent of one another and specify interactions between layers [15]. It facilitates the design and development of the system by enabling system designers to comprehend the operating principles of the system [7, 59–61]. A list of summarized functionalities and tasks of layers in the cited hybrid molecular network model that we are using are described in Table 2.1; meanwhile, Fig. 2.2 explains a subdivision of the physical layer into two sublayers [7]. Higher layer issues and mechanisms in computer networks may apply to MC for application development (i.e., application layer) and reliable end-to-end molecule transmission (i.e., the transport layer) [27] and the relation between the virtual and actual communication at layers in a computer network (from applications to physical layers) [61] also may be utilized in MCs as indicated in [7], in which case is defined a descriptive model and functionalities for each layer (except at application layer) in the hybrid model used in this chapter. This descriptive model establishes molecular:

16

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

Table 2.1 Main functionalities and tasks provided by each layer in molecular communications following the model described by [7] Molecular layer Function 1. Bio-nanomachine location control-environmental monitoring Application 2. Molecule concentration control 3. Structure formation 4. In-network processing 5. Multiscale messaging Transport 1. Molecular transport data unit loss handling 2. Molecular transport data unit flow control 3. Molecular transport data unit congestion control 4. Storing molecular transport data units 1. Network formation Network 2. Routing 3. Molecular packet congestion control 4. Storing molecular packets Link 1. Framing 2. Addressing 3. Molecular frame transmission/reception 4. Molecular frame loss handling 5. Molecular frame flow control 6. Storing molecular frames 1. Signal modulation/demodulation Physical Signaling sublayer 2. Signal molecule transmission/reception 3. Signal propagation/relay/multiplexing 4. Signal molecule error handling 5. Addressing 6. Storing signal molecules 7. Feedback Bio-nanomachine sublayer 1. Acquire and expend energy 2. Replicate molecules 3. Terminate functioning and decompose molecules 4. Move molecules 5. Capture store/release/synthesize molecules 6. Detect/modify molecules 7. Remember/change the state 8. Keep track of time 9. Self-feedback

1. 2. 3. 4.

Communicational components. Communication links. PDUs (Protocol Data Units). Elements required for information storage.

2.1 Molecular Communication Network Architecture

17

Fig. 2.2 A layered architecture of molecular communication

As noted in Figs. 2.1 and 2.2, the upper layers, such as the application and transport layers, may also be included in the architecture of MC. In addition, an architecture of MC may need to consider key issues that cut across layers, such as energy efficiency. An architecture of MC may also need to consider communications at multiple scales in time and space. Finally, standardization of architecture is important for designing a large-scale and complex MC system from bio-nanomachines [8]. The layered architecture described thus far provides advantageous features, such as allowing system designers to focus on issues in one layer by hiding the detail of other layers, similar to typical communication. However, a layered architecture may be violated through a cross-layer design to improve a certain system parameter, such as energy efficiency in MC, quality of service (QoS) (e.g., differentiated communication), or even security. A cross-layer architecture may increase the design complexity but allows a global optimization, which cannot be done with a layered architecture [8]. It is also important to incorporate communications at different scales within the architecture of MC. An MC system may be integrated into the Internet to allow access to bio-nanomachines via the Internet, which could introduce new issues related to micro-scale to macro-scale interaction. An MC system may also be designed hierarchically. For instance, a group of bio-nanomachines may form

18

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

a nanoscale network, a collection of nanoscale networks then forms a microscale network, and interactions may occur across the two very different scales. Biological systems are multiscale, and biological science is developing techniques for multiscale modeling. However, it remains a challenge to determine how the multiscale modeling techniques may apply to extend an architecture of MC or to develop generalized techniques to study the interplay among communications at different scales [8]. Alternative architectures may be adapted from systems biology. One promising architecture is the bow-tie architecture, where interactions among molecules are modeled as a fan-in fan-out network; i.e., a large number of input molecules are converted into a small number of core molecules, which are then converted into a large number of output molecules. For instance, in a cell, energy sources such as metabolites are converted into glucose molecules, which are then converted into many different molecules such as amino acids, sugars, and nucleotides. The bowtie architecture is considered robust since adding a new energy source or removing an existing energy source from the interaction network is simple. Such a network structure identified in systems biology may serve as a foundation for an architectural design of MC systems [8]. The standardization of architecture will be important at some point to facilitate the development of MC systems. In computer networks, standards allow different computers developed by different vendors to communicate. In synthetic biology, a standard library for BioBrick (i.e., bio-nanomachines), called the registry of standard biological parts, is developed for engineering a synthetic living organism from a set of well-defined DNA (deoxyribonucleic acid) sequences. In MC, the IEEE P1906.1 and 1906.1.1 Standards Working Group for Nanonetworking were established in 2011 and 2020, respectively, to promote the standardization of MC. The near-term goal of the working group is to provide a definition of MC, a conceptual framework for MC, and common terminology for MC. The long-term goal is to identify a practical architecture and a set of reusable protocols for MC through the design, implementation, and evaluation of MC systems [8, 34].

2.1.1 Molecular Application Layer As indicated in Fig. 2.1 at the source, the application layer initiates MC by inducing a specific chemical reaction that eventually causes an intended reaction at the destination application layer. The destination reacts to the incoming information molecules produced by the router and initiates an application-dependent action. The application layer provides options to implement application-specific functionality. The specific functionality to be provided by the application layer will become more apparent in the future as applications are being developed [8]. Regarding the descriptive model, [7] mentions that there is not a generic descriptive model for the application layer since it is difficult to create a single descriptive model that applies to many different MC applications due to the diversity

2.1 Molecular Communication Network Architecture

19

of MC applications but effectively, in [7] is provided a set of functionalities for MC applications.

2.1.1.1

Molecular Application Layer Functionalities

A number of fundamental functionalities that the application layer might offer could be useful for many MC applications. Each function at this layer might be backed by features at the physical layer [7]: 1. Bio-nanomachine location control. This functionality is crucial because bio-nanomachines execute application-oriented functions, such as positioning them to illness areas for therapy, which require regulating to establish their distribution across the MC environment. One method of controlling bionanomachine location involves the employment of mobile bio-nanomachines (i.e., bio-nanomachines with “move”) and chemicals that act as attractants or repellents for bio-nanomachines. Either these molecules can be released into the MC environment by bio-nanomachines (with “release molecules”) or they can be pre-stored in molecule storage (with “store molecules”) and injected into the MC environment from an external device, creating gradients of attractants and repellents in the environment. Mobile nanomachines (with “move”) detect concentration gradients (with “modify molecules” and “detect molecules”) and change their locations in response to gradients of attractants/repellants in the surrounding environment [7]. 2. Environmental monitoring. This functionality outlines monitoring the MC environment and identifying a particular event that might happen there. Distributing several bio-nanomachines that can detect a particular signal of interest (with “modify molecules” and “detect molecules”), such as disease-indicating signal molecules released from biological cells, the concentration of particular types of molecules, and environmental disturbances brought on by the intrusion of foreign agents from outside the MC environment, is one method for environmental monitoring [7]. 3. Molecule concentration control. This functionality enables the management of a certain molecule type’s concentration in the MC environment. The usefulness of this capacity is demonstrated, for example, by the ability to keep medication molecules’ concentrations within a certain range. The monitoring of a particular type of molecule’s concentration level by bio-nanomachines using the commands “modify molecules” and “detect molecules” and the command “release molecules” to release molecules that are stored in molecule storage (with the command “store molecules”) when the concentration level of the molecules falls below the threshold is one method for controlling molecular concentration. To regulate the concentration of molecules at particular sites in the MC environment, molecule concentration control and bio-nanomachine placement control may be coupled. Molecular concentration control and bio-nanomachine location control allow for the placement of bio-nanomachines at precise places,

20

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

where the bio-nanomachines use the latter to modify the molecule concentration [7]. 4. Structure formation. The formation of a specific spatial relationship between bio-nanomachines depends on this functionality (e.g., to form a threedimensional structure of a human organ from bio-nanomachines). At the application layer, reaction-diffusion models based on activator and inhibitor molecules may be used to create the required structure. Activator and inhibitor molecules may be stored by bio-nanomachines in molecule storage (with “store molecules”) and released under regulated conditions (with “release molecules”). With the help of “modify molecules” and “detect molecules” and “move,” bio-nanomachines locate quantities of activators and inhibitors produced in the environment and position themselves (depending on the patterns formed spatially) to develop a structure [7]. 5. In-network processing. This functionality enables a network to carry out some computations, such as aggregating environmental conditions detected by a number of bio-nanomachines. Individual bio-nanomachines sense an environmental condition (using “modify molecules” and “detect molecules”), such as a concentration of particular types of molecules, communicate their sensed environmental conditions within a group (using “release molecules” and “capture molecules”), and determine the typical environmental condition as one method of in-network processing (e.g., the average concentration of molecules) [7]. 6. Multiscale messaging. This capability allows information to be transmitted from biological machines to external macro-scale gadgets (like external electrical gadgets) and vice versa. Fluorescence proteins are one method of multiscale messaging. Fluorescent substances can be used in bio-nanomachines so that they respond to the excitation light, which is a message from an external device, and emit fluorescence light, which is a message from the bio-nanomachines to the external device [7].

2.1.2 Molecular Transport Layer The transport layer offers tools for establishing secure end-to-end connections between the source and the destination in order to safely convey data. As a result, this layer may provide services like error handling, flow management, and insequence transmission of information molecules from a source bio-nanomachine to the destination bio-nanomachine, much like what TCP does for the Internet architecture [7, 8].

2.1.2.1

Molecular Transport Layer Communication Model

The molecular transport layer, in contrast to the lower layers, is an end-to-end layer, meaning it only concerns itself with the source and destination of the message and

2.1 Molecular Communication Network Architecture

21

ignores things in between. The fundamental elements of the molecular transport layer are [7]: 1. Molecular transport data unit, PDU. At the molecular transport layer, it is a notion, piece of information, or communication of a statement that both the source and the recipient understand [7]. 2. Source. It sends out a molecular transport PDU. 3. Destination. A source sends a molecular transport data unit, which it receives. A bio-nanomachine, a group of bio-nanomachines, a location, or a collection of various locations in the MC environment are present at the molecular transport layer in an end-to-end MC system (i.e., source or destination) and are typically directly connected to the information source (information destination) [7]. 4. Molecular transport pipe. It is the logical channel through which a molecular transport data unit (PDU) is transmitted between the two communication endpoints (a source and a destination). A network of molecular network layer links known as a molecular transport pipe is used to carry molecular transport data units from one communication end (a source) to the other (a destination). The molecular transport pipe comes in two varieties. The first is a trustworthy pipe that transmits molecular transport data units without error and in chronological order (comparable to TCP functions) [61]). The second pipe is faulty and could transfer molecular transport data units out of order or lose them (similar to how UDP-User Datagram Protocol works) [7, 61]. 5. Molecular transport PDU storage. This logical element serves as a repository for molecular transport data units. The MC environment may serve as the storage location for molecular transport PDUs, while the molecular transport data units diffuse and await processing at their final destination. To store incoming molecular transport data units, a vesicle or liposome lodged in the destination may also be used as a kind of molecular transport PDU storage. When molecular transport data units in the storage deteriorate and lose their functionality over time or when the storage reaches its limit, molecular transport PDUs are lost [7].

2.1.2.2

Molecular Transport Layer Functionalities

1. Molecular transport PDU loss handling. A molecular transport PDU loss handling capability is provided at the molecular transport layer. Loss of a molecular transport PDU at the molecular transport layer can happen either at the destination when the storage capacity for molecular transport data units is at its limit or inside the molecular transport pipe when the MC pipe is congested. Molecular transport data unit loss handling entails three steps: (1) loss detection, (2) loss correction via retransmissions, and (3) molecular transport data unit sequencing [7]. 2. Molecular transport PDU loss detection. For source-initiated loss detection or destination-initiated loss detection to be possible, this feature is necessary (destination-initiated loss detection). Through source-initiated loss detection, a

22

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

source can gauge the strength of the biological reactions that a (lost) molecular transport PDU is supposed to elicit. The degree of the anticipated biochemical reactions that spread through the environment to the sender may not provide an accurate indicator of whether molecular transport data units have successfully reached the receiver because the source and destination at the molecular transport layer can be far apart [7]. 3. Molecular transport PDU retransmission. This capability makes a molecular transport retransmittable. PDU might be started by the destination or the source (source-initiated retransmission) (destination-initiated retransmission). Feedback is necessary for destination-initiated retransmission but not for source-initiated retransmission [7]. 4. Molecular transport PDU sequencing. This functionality is required because a reliable molecular transport pipe requires that molecular transport PDUs be delivered in a specific order to the destination molecular transport layer. For applications like tissue engineering, where molecular transport data units must be given to stem cells in a certain sequence to start a series of the development and differentiation of the stem cells in the appropriate order, sequence control becomes crucial. Following a protocol similar to Stop-and-Wait between the source and the destination (similar to this type of mechanism in digital networks) is an easy method for sequenced delivery of molecular transport PDUs [61]). When a source transmits a molecular transport PDU, it waits for the anticipated biological processes that it is predicted to trigger in the environment to return to the source before transmitting the next molecular transport PDU. The source sends molecular transport PDUs at its pace, and the molecular transport layer at the destination reorders the data units in the storage, which is a more complicated but potentially more effective solution. For this reason, every molecular transport PDU needs a special sequence number. Additionally, the storage of molecular transport PDUs must be able to be sorted by the molecular transport layer, or the molecular transport data storage must be able to transfer its contents sequentially [7]. 5. Molecular transport PDU flow control. This functionality allows a source to modify the rate at which molecular transport data units are transmitted in order to prevent loss at a destination. This capability and the flow control at the molecular link layer are essentially identical [7]: • Source-initiated flow control. To prevent any molecular transport data units from being lost at the destination, a source may employ a very low transmission rate. As an alternative, a source may use “molecular transport data unit loss detection” to detect loss and change the transmission rate [7]. • Destination-initiated flow control. The source adjusts its transmission rate in response to the size of the available molecular transport data unit storage at the destination as a destination’s molecular transport data storage begins to fill up [7]. The conventional tiered architecture of communication networks is where this type of flow control is most prevalent. In order to do this, the destination

2.1 Molecular Communication Network Architecture

23

must gauge the amount of storage space available for molecular transport data and report back to the source [7]. 6. Molecular transport data unit congestion control. To prevent congestion, there is a functionality to control the quantity of molecular transport data units in the molecular transport pipe. Controlling the source’s transmission rate can help to limit the amount of molecular transport data units [7]. A source may decrease the rate at which it transmits molecular transport data units after discovering a loss through “molecular transport data unit loss detection.” If the constant value is stored in a simple memory, the transmission rate can be decreased to a predetermined and constant pace. The rate may also be decreased to a value that is dependent on the level of congestion, necessitating the capacity to both detect the level of congestion (e.g., through “molecular transport data unit loss detection”) and determine the new transmission rate in accordance with the level of congestion [7]. In order to fully utilize the capacity of the molecular transport pipe, the source may start increasing the transmission rate whenever the rate is reduced after loss is detected. It is possible to raise the transmission rate either by a fixed amount, which simply requires a simple memory, or by a variable amount that is dependent on factors like how long it has been since a loss occurred, which necessitates the ability to recall the past [7]. 7. Storing molecular transport data units. In the molecular transport data unit storage, a source or destination keeps molecular transport data units [7].

2.1.2.3

Significant Cases of Investigations in Molecular Transport Layer

By exclusively releasing medication molecules close to the target site, specific transmission control protocol (TCP) features are used in cellular communication to dispense pharmaceuticals. This lowers adverse effects and increases treatment effectiveness. To initiate and terminate the treatment, a connection-oriented protocol locates a receiver (such as tumor cells) and initiates and terminates communication. In order to prevent receiver congestion, a proper transmission rate (flow control) is utilized between the transmitter and receiver. Additionally, dependable data transport guarantees that the desired number of molecules is obtained. When all of the molecules have been delivered, the receiver sends a stop signal to cut off the connection and cease the medication flow. At the receiver, flow control is also applied. When the receiver is full, a method is used to prevent the transmission of medication molecules [15, 62, 63]. In [64] is presented a specialized transport layer protocol with a delicate balance among throughput, delivery ratio, energy consumption, and end-to-end delay for MCs in BANs (body area nanonetworks) [65]. In order to achieve the desired performance level, this transport protocol uses a notion of conjugate congestion control from the viewpoints of both the sender and the receiver. It also produces minimal complexity and overhead (in contrast to the significant overhead produced

24

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

by TCP in a network) through congestion-control-based transport layer protocol, leaving the other structures closer to the conventional protocol UDP. The reliability of MC is then increased by causing acknowledgments for molecules that are successfully received at the receiver’s end. The suggested protocol provides a balance between several network-level performance parameters while requiring little overhead and complexity from BANs, ensuring allowable throughput and an acceptable average end-to-end delay, making it suitable for a variety of BAN applications. Multiple senders causing congestion and the receiver’s inability to handle the enormous number of molecules could lead to data loss, serious mistakes, and subpar performance. Power consumption and data processing speed during communication are major problems because of the nanoscale size of the nanomachines. So that the transmitter’s current molecular rate always stays within the range of the receiver’s processing capacity, the receiver nanomachine must give some kind of acknowledgment to the sender nanomachine. In order to continuously adjust the molecule injection rate during communication, the disclosed transport layer protocol includes a conjugate congestion-controlled acknowledgment-based technique that involves both sender and recipient nanomachines [64]. Fundamentally, the protocol defines actions that must be taken at the transmitter and receiver ends to complete acceptable transport layer communication. A data molecule is then sent and a timer is started by the sender nanomachine. The countdown is stopped if the sender receives a confirmation from the recipient prior to the timeout. Then, suppose the receiving amount matches the sending amount or is set to equal the receiving amount. In that case, it doubles M, which stands for the number of molecules, according to how many molecules were successfully received by the receiver. After sending this modified M number of molecules, it restarts the timer. However, the sender reduces the amount of molecules M and starts a timer after transmitting M molecules if it does not get the acknowledgment inside the timer. The N variable, which represents the anticipated number of molecules to be received, is maintained by the receiver nanomachine and is continuously changed in response to the data molecules received. It begins a timer after getting the first molecule and waits until it receives N molecules. Prior to timeout, if it receives N molecules, an acknowledgment is delivered, and the value of N is raised. To warn the sender about potential data loss in the event of a timeout, the receiver sends an acknowledgment that includes the value of N. The value of N is subsequently reduced [64]. Energy efficiency is ensured with this transport protocol by the absence of continuous acknowledgment and retransmission. Additionally, sporadically receiving an acknowledgment from the recipient gives the sender the ability to choose how much data to send next, which limits the likelihood of network congestion. The flow of the congestion-controlled acknowledgment-based transport layer protocol that we provide is shown in Fig. 2.3. There are four parameters in this technique that require adjusting. These are the timers Ts, Tr, expected number of molecules N, and molecule injection rate M. The other two are on the receiver side, where as M and Ts are on the transmitter side. Depending on the acknowledgment, M is either

2.1 Molecular Communication Network Architecture

25

Fig. 2.3 Congestion-controlled transport layer protocol

doubled or cut in half. If N successfully gets the predicted number of molecules, N is doubled; otherwise, N is left intact. In order for it to wait based on the amount of sent and anticipated molecules, Ts and Tr are configured to be multiples of M and N [64]. In [66], for complex real-time healthcare applications, a hybrid communication system that combines molecular and electromagnetic communication is envisaged. Therefore, in this hybrid system, BAN communication is seen as a tiered protocol stack, similar to TCP/IP for MC. The Channel Response Quantum model is the physical layer’s propagation model used by the authors of [66]. This model uses a diffusion process to propagate molecules and adheres to Fick’s law. This model predicts the values for the received molecule concentration per unit volume as follows [67–69]: .

U (r, t) =

Q 4π Dt

−r 2

e 4Dt

(2.1)

where r, Q, and D are the receiver’s distance from the emitting source, the amount of released molecules per second, and the diffusion constant in .cm2 /s unit, respectively. During the transmission of a message in MC, the message is considered a 1-bit if more than a specific number of molecules are received during a period .tH . Otherwise, it is considered as 0.

26

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

The network can get readily crowded because BANs’ healthcare applications call for sending significant volumes of data among the nanomachines. The receiver nanomachines then fail to process the majority of molecules when more data are added to a busy network. As a result, the network’s overall performance tends to suffer. As a result, a transport layer protocol with a congestion control mechanism is created, in which the receiver alerts the sender nanomachines to lower the data rate in the event of congestion. The molecular nanonodes and the electromagnetic nanonodes in this model of a hybrid nanonetwork must communicate with one another. They cannot directly interchange data or information because they are two entirely different types of entities, and their communication methods and functions are also different. Therefore, a gateway must be installed between them to perform this function for them. So, a hybrid nanonetwork consists of three different types of nanonodes: the molecular node (M), the electromagnetic node (E), and the gateway node (G). When a node (M) needs to communicate information to a node (E), it first sends that information to a node (G), and node (G) then transmits that information to the appropriate node (E). When routing, the network layer handles data transmission via a gateway node. Data transfer from a molecular node is proposed as a transport layer protocol using hybrid communication to improve the performance of BANs while maintaining the other structures in line with traditional UDP [66]. For communication between nanomachines in BANs, power consumption is a major problem. Due to the nanomachines’ small size and inability to process huge amounts of molecules, congestion brought on by numerous senders may cause data loss and performance degradation. Because of this, effective communication in BANs requires the sender nanomachine to receive some acknowledgment in order to determine whether the receiver can handle the present rate of molecules. Given this, the transport layer protocol controls a system of congestion to adjust the molecule injection rate during communication by instructing the sender to wait for a certain amount of time before transmitting the subsequent data [66].

2.1.3 Molecular Network Layer This layer provides methods for a collection of bio-nanomachines to exchange information over distances greater than those attained by the link layer. The network layer’s mechanisms include distributing bio-nanomachines throughout the environment to create a network of machines connected by communication links (channels), choosing communication channels for machines to use to transmit information from a sender machine to a receiver machine (network routing), processing data within a network of machines, and managing congestion in a network of machines [7, 8, 70].

2.1 Molecular Communication Network Architecture

2.1.3.1

27

Molecular Network Layer Communication Model

The following are the fundamental elements of the molecular network layer, which are identical to those of the molecular link layer [7, 71]: • Molecular packet. At the molecular network layer, it refers to an idea, a fact, a communication, or an assertion that is understood by the source and the destination as well as the molecular router [7]. • Source. It broadcasts a packet of molecules. A bio-nanomachine, a collection of bio-nanomachines, or a site that is typically directly linked to the information source are all examples of sources [7]. • Destination. A source sends a molecular packet, which it receives. A destination is typically directly connected to the information destination and can be a bionanomachine, a collection of bio-nanomachines, a place, or a collection of several locations in the environment [7]. • Molecular communication network. The molecular packet is sent from a source to a destination across a set of molecular network links and routers. An MC network could be as basic as the area in the environment where molecular packets randomly spread from a source to a destination at the physical layer. It might be more complicated and comprise, for example, a number of protein filaments producing a complex network topology linking all potential sources and destinations [7]. • Molecular router. It belongs to an MC network as a logical element. It examines a molecular packet’s destination address information before forwarding the molecular packet to one of its outgoing molecular network links in the direction of the destination. In the MC environment, where molecule packets can easily spread and diffuse, a molecular router can be constructed upon a bio-nanomachine. An arriving molecular packet is forwarded directionally via a molecular router in the direction of the destination. A bio-nanomachine that connects sources and destinations on a network of protein filaments may also be referred to as a molecular router. A molecular router sets up circumstances so that an approaching protein motor carrying a molecule packet is forced to take a protein filament that leads to the destination when it reaches the intersection [7]. • Molecular network link. It serves as the logical conduit for molecular packet transmission between two nearby molecular network nodes (sources, destinations, and molecular routers). Multiple MC linkages linking the same two molecular network nodes at the molecular link layer may make up a single molecular network link connecting two molecular network nodes at the molecular network layer [7]. At the molecular network layer, a molecular network node may be connected to several molecular network links. A molecular network node may directionally release a molecular packet (and, as a result, the molecular packet directionally propagates through the environment) if, at the physical layer, an MC network is supported by the area in the environment where molecular

28

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

packets propagate stochastically from a source to a destination, giving a choice in the directions (i.e., molecular network links) to forward a molecular packet toward. A molecular network node may attach several protein filaments if an MC network at the physical layer is made up of various protein filaments that create a network topology, providing a selection of molecular network links to send a molecular packet onto [7]. If a protein motor carrying a molecular packet has internal capabilities to decide which crossing protein filament to take to reach its destination, the molecular router may also be a protein filament intersection on a network of protein filaments. An incoming protein motor here serves as a routing function [7]. • Molecular packet storage. The storage of molecular packets is a logical component at a molecular network node. When a molecular packet in a storage device degrades and ceases to operate as a molecular packet, or when the storage device’s capacity is reached, molecular packets are lost [7].

2.1.3.2

Molecular Network Layer Functionalities

1. Network formation. Between sources and destinations, there is a capability to create an MC network (i.e., between source and destination bio-nanomachines or between source and destination locations at the physical layer). From source to destination, molecular packets are sent through the network created by the network construction functionality [7]. Physical layer functions may help networks form at the molecular network layer. Using a self-organizing protein network is one method for creating networks. Bio-nanomachines coupled to particular molecules, or “seed” molecules, produce protein filaments, or microtubules, on their own and connect to other bio-nanomachines through dynamic instability. Once these connections are made between bio-nanomachines, protein motors travel along the interconnections, transporting signal molecules from one place to another. Extending an addressable space at the physical layer, where addressable spaces based on concentration gradients of guide molecules may be concatenated to construct an MC network, is another physical layer technique for network development [7]. 2. Routing. It has the ability to regulate the routes that molecular packets take to get from one place to another [7]. Functionalities at the physical layer may facilitate routing at the molecular network layer. Utilizing the stochastic behavior of protein motors at the intersections of protein filaments on an MC network is one method at the physical layer to assist routing at the molecular network layer. A protein motor transporting molecular packets (or signal molecules) may choose to move along the same protein filament or switch to a crossing protein filament at a protein filament intersection. It may be possible to control the stochastic behavior of protein

2.1 Molecular Communication Network Architecture

29

motors at protein filament intersections, guiding protein motors to the location where molecular packets (or signal molecules) are delivered, by adjusting factors like the size of protein motors, the size of protein filament links, and the distance between the crossing protein filament links [7]. By adding a router bio-nanomachine that directs incoming molecular packets to other bio-nanomachines that are headed in the direction of the destination, another method at the physical layer to allow routing at the molecular network layer can be achieved [7]. 3. Molecular packet congestion control. To prevent congestion in the MC network, there is a functionality that controls the volume of molecular packets. One way to reduce the number of molecular packets is to either discard them at a molecular router or control the rate at which they join the MC network [8]. • Molecular packet discarding. A molecular router may discard new molecular packets or certain molecular packets that are already in the molecular packet storage when the molecular packet storage is at or near its limit in order to prevent congestion in the MC network. When the molecular storage reaches its limit, the simplest and most natural discarding mechanism is to stop accepting new molecular packets. To distinguish between different molecular packets in the storage, this does not need a molecular network layer. Due to its dependence on underlying physiological reactions of the molecule storage, this simplest packet discarding is easily supported by features at the physical layer (e.g., biochemical reactions that govern vesicles intaking of molecules) [7]. • Molecular packet rate control. The number of molecular packets that enter the MC network may be controlled by a molecular router. It is possible to predetermine the rate of molecular packets or to find out through feedback from other molecular routers. By adding a specified time lag between two successive molecular packets, a molecular router, for instance, may forward molecular packets at a consistent rate. As an alternative, a molecular router may change the interval between two subsequent molecular packets in response to feedback (or backpressure) from other molecular routers in the area. Physical layer functionalities like “keep track of time” and “release molecules” can provide molecular packet rate management utilizing the predetermined rate without the need for feedback (to insert time lag between successive molecular packet transmissions). The molecular network layer may tell the physical layer of the frequency of molecular packet transmission. Feedback is necessary for molecular packet rate control that modifies the rate [7]. 4. Storing molecular packets. A molecular network node holds molecular packets in the molecular packet storage, such as a source, destination, or molecular router [7].

30

2.1.3.3

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

Significant Research Studies in Molecular Network Layer

A molecular packet, according to [72], has a structure that is remarkably similar to an Internet packet, consisting of a header and a payload. The Internet Protocol (IP), which divides an IP datagram into smaller fragments to ensure that the fragments can transit across underlying communication channels, served as inspiration for the fundamental concept suggested in [72]. To make it easier for a huge molecular message to diffuse along a transmission drift channel and reach its biological target, it is divided into smaller packets [73]. Furthermore, various molecular packets are transmitted using the channel’s linear features [20]. The receiver at the destination decodes a group of fragments and puts the molecular message back together. Also, since IP-based communication networks are similar to them, more researchers with experience in Internet infrastructure networks should be drawn to this field. Vesicle transmission through nanotubes would resemble the movement of packets through a network link. IP packet flows and bacterial and viral gliding might both be compared for connected nodes. However, bacteria and viruses are not found inside the tubes. Research opportunities include multipath routing, route and path managements along tunneling nanotube networks, and resource management along tubular linkages [16]. Because the principles of molecular routing are covered in [27], the authors prove that routing ideas from computer networks can be used to make MC networks larger and more complicated. The range of MC is constrained by the maximum distance that molecules may propagate with reliability in the absence of routing. When traveling farther, molecules become dispersed and are either too weakly concentrated to be recognized by the receiver bio-nanomachine or are distributed and unlikely to arrive at the machine in an acceptable period of time. Routing in MC may involve router bio-nanomachines situated along the route to the receiver bio-nanomachine, just like in computer networks. MC is currently restricted to static routing tables, which are inflexible to changing network or locational conditions. For instance, in [74] and [27] a sender bio-nanomachine communicates data using a bacterium that has addressing molecules (such as type X and Y molecules) to identify the desired receiver bio-nanomachines. This technology uses a bacterium that travels along chemical gradients to the next router bio-nanomachine before transmitting information using statically set chemical processes to the receiver bio-nanomachine. To develop a routing system based on bacteria, a comparable mechanism is being investigated in [27, 75]. How to create a dynamic routing system that can adapt to dynamic situations, such as bionanomachines moving around the environment dynamically, is a problem that needs to be addressed [27]. Even when the present MCs’ routing is restricted to static routing tables, this is insufficient to account for the network nodes’ dynamic positions. Static routing involves a sender transmitting data using a bacterium that has addressing molecules that identify the intended recipient. A bacterium that follows chemical gradients is received by the router node, and it uses statically programmed chemical processes to retransmit the data to the next target router [25].

2.1 Molecular Communication Network Architecture

31

Fig. 2.4 Molecular communication and network model

The proposed system model in [76] which uses a cooperative communication system with two-hop communication is illustrated in Fig. 2.4. (each node is used for signal repeating to improve system reliability). More relay nanomachines will increase the number of hops. The unguided diffusion-based MC is employed in each hop to make wireless communication between nanomachines possible. The system consists of elements acting as information molecules that represent the data to be transmitted, transmitter nanomachines that emit the information molecules, receiver nanomachines that detect the information molecules, relay nanomachines that can emit and receive molecules, and the environment in which the information molecules travel from the transmitter nanomachine to the receiver nanomachine. In order to offer an addressing method, it is assumed that each nanomachine contains a self-identifying label that may be affixed to the molecules conveyed. The sender address, information, and receiver address are all contained in the information molecule. Each hop in the communication process involves the following steps: the transmitter nanomachine encoding information into an information molecule, the transmitter nanomachine transmitting the information molecule into the environment, the transmitter nanomachine propagating the information molecule through the environment, the receiver nanomachine receiving the information molecule, and the receiver nanomachine decoding the information molecule into a chemical reaction [76, 77]. To carry out organic processes, a certain type of “addressing” is used throughout nature. Thus, a gene is described in biology as a collection of nucleotides that contains the data necessary for a crucial function to be carried out at the target,

32

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

which can be done by a protein or ribonucleic acid (RNA Ribonucleic acid). The processing of this biological information is understood by a specific target organ (to carry out a biological function) at the destination through intercellular communication, and as a result, the contents of a specific gene may be associated with a network layer address in terms of networks (long-distance) [15]. For the application interface stack, which encrypts messages and addresses; the network stack, which routes communications; and the error correction stack, which encrypts and decrypts error correction codes, hybrid DNA-based computations were proposed in [12]. The linkswitching stack was suggested to use enzyme-based computation. It was expected that DNA base pairs, which behave similarly to the network layer in conventional networks, would be utilized for addressing and transporting information in place of bits. In [78] an MC nanonetwork design with two nodes, a nanomachine and a nanosink, is presented. The fundamental unit for completing activities at the nanoscale, nanomachines can only convey information across small distances (nm-.μm). Nanosinks are a distinct class of nanomachines with larger sizes and more potent processing and communication capacities. A nanosink, numerous nanomachines, and a sub-nanonetwork are depicted in Fig. 2.5. Using multi-hop methods, the nanomachines will communicate with the nanosink. A nanonetwork is connected to several sub-nanonetworks by medium-range MC (.μm-mm) between nanosinks. The free-diffusion-based MCs are an efficient short-range communication technology that allows information molecules released from nanomachines to

Fig. 2.5 An architecture of nanonetwork utilized in [78]

2.1 Molecular Communication Network Architecture

33

diffuse in any direction over relatively short distances without the need for additional communication infrastructure. The concentration level and rate of concentration change can be employed as the information carrier in MCs, similar to amplitude modulating (AM) and frequency modulating (FM) technologies in conventional communication, such as electromagnetic waves [78]. One sub-nanonetwork of the nanonetwork is used to transmit information via the free diffusion-based MC from the source (nanomachine) to the destination (nanosink). In a nanonetwork, the sender diffuses information molecules into a propagation medium, receiver nanomachines2 are outfitted with the precise kind of receptors, and only when the concentration of the right kind of information molecules reaches a certain level does it begin to receive information molecules. Given that a nanosink is larger than a nanomachine, it can emit a significant amount of the particular molecules whose concentration gradient serves as an “indicator” to show which nanomachine is located closer to the nanosink, which can then be used to decide the next hop for data relay [78]. The three categories of in-body, on-body, and off-body in nanonetworks are analogous to those in conventional body-centric communication. Figure 2.6 depicts the general structure of a nanonetwork for the healthcare industry. It can be summed up as follows [6]: • Nanonodes. These nanodevices are the smallest and most basic. They can only carry out rudimentary computation tasks and communicate over very short distances because of the low energy, memory, and communication capabilities. The nodes could have communication and sensing components [78]. • Nanorouters. These nanodevices, which have slightly more processing power than nanonodes, may collect data from a small number of nanomachines and regulate the actions of nanonodes by transmitting really basic commands (such as on/off, sleep, read value, etc.). But since they would get larger as a result, their deployment would be more intrusive [6, 78]. • Nano-micro interface. They are utilized to gather the data sent by nanorouters and transfer it to the micro-scale devices. They can simultaneously transmit data from the nano-scale to the micro-scale. Nano-micro interfaces are hybrid devices that can use both traditional communication paradigms and nano-communication approaches to communicate in micro/macro communication networks [78]. • Gateway. It enables users to remotely control or keep an eye on the complete system over the Internet [78]. The interest in the development of nanodevices, their interconnectivity to realize the Internet-of-nano-things (IoNT), and their applicability in biomedical applications is rising quickly with the advent of novel nanomaterials (e.g., graphene-based

2 We will use the term nanomachine or bio-nanomachine according as it is gotten from the original papers that provide this information. Besides, the only difference between terms nanomachine and bio-nanomachine is that bio-nanomachine is a nano-to-micro-scale device composed of biological materials and capable of simple chemical functions [7].

34

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

Fig. 2.6 Envisioned architecture for nano-healthcare

carbon nanotubes) with remarkable properties. The wireless-capable nanoscale devices are less intrusive than conventional techniques, such as endoscopy, making them an excellent choice for in vivo biomedical applications. Because of its nonionization and resistance to fading properties, the terahertz (THz) band (0.1–10 THz) is regarded as the most promising band for the functioning of nanodevices. The envisioned structure for in vivo nanonetwork healthcare is provided in [79, 80], and is depicted in Fig. 2.6. In these circumstances, the nanodevices must possess the following qualities: the system can be operated remotely over a WAN or the Internet, thanks to a gateway or access point (AP), which collects data from a large number of nanonodes and can send a limited number of commands. Nanonodes, which are composed of sensor and communication units, can transmit over short distances to generate simple computations. Nanonetworks can connect nanomachines to increase the restricted communication range, allowing for cooperation and information sharing as well as facilitating the connectivity of nanomachines to the Internet [81, 82]. IoBNT [83, 84], a version of IoNT, is envisioned as a heterogeneous network of nanoscale bioelectronic components and modified biological cells, referred to as Bio-NanoThings (BNT), which may communicate via electromagnetic waves and via MC. The goal of this notion is to work directly with the cells, allowing for more precise sensing and, eventually, real-time control of the complex biological dynamics of the human body. IoBNT’s strategy necessitates that the modified cells sense, process, and communicate with one another as well as with external devices that offer remote and least invasive methods of interrogation. The invention of implanted submillimeter BNTs that can sense biochemical data in the human body and communicate that data remotely to a wearable hub outside the body is the first step in the realization of IoBNT [85].

2.1 Molecular Communication Network Architecture

35

By allowing in vivo continuous monitoring of infections by implanted nanoscale sensors detecting communication among infectious organisms within the body, the IoBNT framework distinguishes itself from current existing technologies. These sensors then transmit data to a wearable mobile hub, which sends the gathered information to medical specialists. As a result, the patient does not need to go to a lab for testing, and infections can be identified early, even before symptoms show up, leading the patient to seek medical help. In this manner, the possibility of susceptible patients dying too soon can be decreased [85]. Particularly for cancer patients who are immunosuppressed after chemotherapy and susceptible to significant infections, which is a major cause of mortality, early diagnosis of infections is essential. Another illustration would be infections, which happen wave by wave and result in the patient’s death in the case of cystic fibrosis, a genetic illness with no known treatment that mostly affects the lungs. The quality of life and life expectancy of cystic fibrosis patients will therefore be enhanced by early identification of lung infections. Moreover, early infection detection in patient populations who are at risk will enable prompt administration of antibiotics and other medications, minimize hospital stays for treatment, reduce mortality, and result in a considerable drop in healthcare costs. Also, treating infections is getting more and more difficult for medical personnel due to the growth of antibiotic resistance among infectious germs [85]. The improper antibiotic being used can prolong therapy and drastically lower survival rates. This IoBNT program can also be used to monitor the effectiveness of medications. This system promotes both public health and the health of individuals. The ongoing infection monitoring that IoBNT systems offer is highly beneficial in the event of an epidemic or pandemic. IoBNT can be readily customized for tracking, tracing, and quarantining persons, especially since they are already coupled with mobile devices and remote data analytics systems [85]. The environment shown in Fig. 2.6 also represents inter-nanomachine communication for medical applications and proposes a new paradigm for BANs. By using biocompatible molecules as communication carriers, MC enables nanomachines to communicate with one another, making it a viable alternative for BANs. The development of nanotechnology opens the door for the creation of nanomachines that can carry out extremely straightforward and narrowly focused functions at the nanoscale, such as computation, data storage, sensing, and actuation for a variety of applications, such as in the biomedical industry. Nanomachines can be created to carry out extremely complex tasks, such as identifying and eliminating cancerous cells by penetrating delicate bodily areas (e.g., the spinal cord and gastrointestinal, among others). Promoting nanomachines’ ability to communicate with one another in order to make judgments quickly in order to cure complex diseases is crucial to doing this [16, 66, 86]. An essential IoBNT use through mobile bio-nanomachines in the blood capillary and surrounding tissue is examined in [87]. The transmitting bio-nanomachine is described as a rotating sphere with a steady stream of emissions for a predetermined amount of time. In order to assess their impact on the molecular received signal, the blood capillary properties, including the blood-tissue barrier and blood flow,

36

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

are modeled. The impact of the emission duration, elimination rate, and separation distance on the molecule received signal is the main subject of that investigation. Due to its strong characteristics, including biocompatibility, low energy needs, and compact size, the basic transmission of MCs is a more practical and effective communication technology between nanomachines inside the body (i.e., for IoBNT applications) [88]. The capacity of nanorobots to conduct controlled tasks (such drug release) might then be achieved at a laboratory scale, i.e., a DNA-based nanorobot, since it is not constrained by antenna size and frequency requirements. A group of bio-nanomachines that are moving through the blood and surrounding tissues form an intra-body nanonetwork in Fig. 2.7, where the strategy used in [87] is implemented. In this figure, they collaborate to perform specific tasks like sensing the biological and chemical changes in the human body. The intra-body nanonetworks can also activate medical devices using commands supplied remotely from an outside device. Future nanomachines are anticipated to communicate with one another via MC in order to collaborate on a variety of challenging sensing and actuation tasks. A bio-cyber interface can be used to connect the intra-body nanonetwork-connected communicative nanomachines to an outside network (such as the Internet) in order to increase their efficiency and facilitate more intricate real-world biomedical applications. The bio-cyber interface transforms intra-body nanonetwork chemical impulses into electrical signals for the Internet’s cyber realm (and vice versa). Consequently, the Internet of Bio-Nano-Things (IoBNT) framework, depicted in Fig. 2.7, will be created by the communication between the nanomachines and the external Internet network via the bio-cyber interface. The IoBNT can allow internal nanomachines to transmit acquired biosensor data to an outside monitoring device for additional processing over the Internet. The nanomachines can also be remotely controlled by an outside party to carry out specified activities like drug release and medical actuation. In this case, it is believed that the bio-cyber interface will connect to a nearby gateway device (like

Fig. 2.7 Typical Internet of Bio-Nano-Things (IoBNT) used in [89]

2.1 Molecular Communication Network Architecture

37

a smartphone) and communicate with it over a short distance utilizing a lowpower communication method like Bluetooth Low Energy or Radio-Frequency Identification. The data can then be transmitted and received by the smartphone using a remote access point linked to the Internet network. As an alternative, the bio-cyber interface might be made to visualize the biochemical data using flexible skin-attachable screens, such as light-emitting diodes and capacitors. After that, the captured visual data can be read out on a smartphone and sent to the Internet network through the remote access point [87]. It is generally accepted that MC is one of the most practical communication technologies to implement BANs for IoBNT, allowing for a large reduction in compatibility and potential harm to the body area to a minimum level. To accomplish this, a network topology based on MC must be established, in which communication nodes must be statically deployed at some important human body locations (such as the pancreas) to carry out their functions (such as to detect early diabetes), and all communication parameters must be controlled in accordance with the predefined topology to achieve the lowest possible network error. The implementation of associated body area network experiments should be based mostly on biomedicine since it is important to consider human experimentation’s security and dependability carefully. However conducting this kind of real network experiment still presents a significant problem due to the limitations imposed by hardware designs and laws [90]. Compared to conventional electromagnetic wave-based communication, MC is a cross-disciplinary research area involving several fields, including biology, computer science, communication engineering, and more. It exhibits notable properties. First of all, MC uses less energy and increases energy effectiveness. The chemical reactions taking place in the body region environment can provide MC with enough energy to allow information transfer. For instance, the informationcarrying tool myosin may utilize chemical energy for mechanical actions with nearly 100% efficiency. Second, since MC uses molecules as the information carrier, a lot of wireless communication is projected to be ineffective or impracticable (e.g., nano-sensor networks based on MC can detect micro bacteria or toxic substances that cannot be detected by traditional sensing technology). Finally, MC is suited for the body’s nano-network region and is biocompatible. For instance, MCbased nanonetworks can offer the medical industry minimally intrusive medical technologies. Due to their unique characteristics, MCs have drawn a lot of attention since they are viewed as an alternative to electromagnetic communications [74]. MC is less suited for real-time healthcare applications since it is typically slower and more prone to errors than electromagnetic transmission. The abovementioned problems pertaining to complex real-time healthcare applications are resolved by a hybrid communication that combines electromagnetic and molecular communication [74]. Thus, Fig. 2.8 illustrates the topology used in [74] for this setting. Molecular node (M), electromagnetic node (E), and gateway node are the three different types of nanonodes that make up this hybrid nanonetwork (G). When node M wants to communicate information to node E, it first sends that information to node G, which

38

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

Fig. 2.8 TRouting in heterogeneous communications of BANs

then sends it on to the appropriate node E. While routing, the network layer handles data transmission via a gateway node, the data transmission across a gateway from a molecular node to an electromagnetic node [74]. In order to prevent electromagnetic radiation from reaching sensitive body parts while maintaining a low error rate and short delay for nano-communication-based BANs, a hybrid nanonetwork can be employed as a solution. The suggested routing protocol in [90] for the aforementioned study makes sure that data is sent from the molecular node to the gateway, which then delivers data to the electromagnetic node. Only the molecular nodes around node M are added to the path while looking for a way to get from molecular node M to gateway node G. Building a path from node G to electromagnetic node E further is accomplished in a similar manner. Only nodes of the same type and the gateway are included in the routing table and neighbor list of a node for this reason. All types of nodes are listed in the neighbor list and routing table of a gateway node. In order to find a path from node M to node E, it naturally finds a route through the gateway. The authors in [90] proposed a nano-communication system in the preceding BAN that minimizes overhead (comparable to traditional UDP), offers enough throughput and end-to-end latency, uses less power, and manages congestion of body area nanonetworks. The receiver tells the sender of nanomachines to slow the data rate (molecule injection) in case of congestion in this transport layer’s congestion control mechanism. It is a suitable design approach for applications in the healthcare industry [90]. A unique routing strategy that makes use of a concentration gradient to direct routing toward the sink is put forth in [91]. They create an easy-to-implement

2.1 Molecular Communication Network Architecture

39

routing method to help information travel toward a sink node by drawing on bacterial features including directional sensing and molecular prioritization. The features of a multimolecular field are explored in simulation research by varying factors like the rebroadcast period, the relay nodes’ retransmission threshold, and the transmission strength. The effectiveness of the strategy is also examined in terms of the likelihood of successful message delivery and the resulting propagation delay. Findings show that short propagation delays can be maintained while achieving successful delivery probabilities close to 1. This is possible with the right selection of design variables. According to [91, 92], a routing system using virus particles as information carriers–with the information being encoded in the DNA or RNA component of the viral particle–was presented. According to numerical calculations, adding more intermediary nodes increased the planned network’s reliability at the expense of a longer overall transmission time. In contrast, specified pathways were taken into account for multipath routing by addressing the transmitted virus through the envelope gene and using the proper receptors at the recipient solely to receive the desired information. The research in [91, 93], and [94] which uses bacteria as information carriers based on the early work of [74, 91] is another intriguing routing approach. While using numerous attractant molecules to construct a virtual network topology for multipath communication, a single attractant was used to represent a single link. Later, this approach was modified in [91, 94] to imitate IP-based communication networks’ opportunistic routing, packet addressing, and packet filtering processes. Another intriguing strategy is that described in [91, 95], where an opportunistic routing protocol is suggested to get around the drawbacks of the addressing strategies looked at in earlier strategies. The implementation includes mechanisms for nearby node identification, prioritizing based on the distance to the source node, and acknowledgments for coordination. The method depends on the creation of a gradient field to direct the transmission toward the gateway. The authors of [96] describe a sensoring network for a biological or medical application in which a hybrid strategy combining molecular motors and diffusion is used for communications at a short-range level and a bacteria communication strategy is used to establish communication at a medium-range level. The various range levels are united using the same DNA encoding technique. Certain characteristics of bacteria-based communication, such as the capacity to load several plasmids, or DNA-encoded messages, onto a single microbe, make it suitable for combining data from multiple sources and conveying a larger amount of data at once. A communication produced at the short-range level, which consists of tiny nanomachines and nanosensors, can be guided across the network until it reaches a nanogateway or nanointerface that might link to, for example, a personal area network that is connected to the Internet. This approach is thought to allow for the individual targeting of nanogateways at the medium-range level employing various chemoattractant [96]. The suggested fix for the aforementioned system causes the little nanomachines and nanosensors to take up a position. If the releasing point is located in the apex of

40

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

Fig. 2.9 Example DNA packet encoded into the carrier plasmid

the nanomachine, the transmission will always be upward because mono-monostatic body nanomachines or nanomachines with a gömböc3 integrated would force them to always stand in the same posture. The density of the nanomachines can be changed such that some are “heavier” and sink in the surroundings while others “float in the middle.” As a result, the basic nanomachines, or nanosensors, would sink and only broadcast upward, while the gateway nanomachines would hover nearby, take in the data, and use bacteria to interact with one another [96]. The packet that is envisioned for the system in question is relatively straightforward, consisting of three information blocks: two for the source and destination and one for the actual information. There are three blocks for source and destination. Since one block targets the destination nanomachine and the other the destination nanogateway, one block is needed for the source and two for the destination. It is crucial to distinguish the message from other data in the plasmid since the message is stored in a series of base pairs and there are several blocks of information. A method that manipulates various restriction enzymes must be used in the system in order to fulfill this communication requirement. An enzyme known as a restriction enzyme recognizes a specific base pair sequence and cuts DNA at the restriction site. To distinguish all blocks of information in this system concept from other information in the plasmid, two restriction enzymes are used to wrap them, as shown in Fig. 2.9. After removing the message from the plasmid, a nanogateway can scan the header blocks to determine where to direct the message [96]. The mechanism regulates distinct chemoattractants to address various nanogateways and route the information accordingly. Each nanogateway has the capacity to produce bacteria that are selective and will only respond to a certain chemoattractant. These chemoattractants can be converted into value addresses; for instance, if a nanogateway has the address AATC, other nanogateways will use bacteria that will respond to the AATC chemoattractant to communicate with it [96].

3 A gömböc is a mono-monostatic body, which is a convex three-dimensional homogeneous body, which when resting in a surface has one stable and unstable point of equilibrium.

2.1 Molecular Communication Network Architecture

41

It is necessary to have a translation table or a process that identifies what kind of bacterium should be used to respond to a given address. The nanomachines will search for a nanogateway when the nanonetwork is deployed, associating with the first nanogateway they locate. When a nanomachine sends a request to the nanogateway, the nanogateway will begin a process in which it will assign an address for that nanomachine based on its address and respond to the nanomachine with a message indicating that it has acknowledged the request and including the address of the nanomachine. The closest nanogateway or the closest nanogateway that could accommodate his request will be the one the nanomachine associates with based on whose nanogateway’s response arrived first. A single nanogateway may be connected to many nanomachines, and when the nanogateway later communicates information by diffusing molecules downhill, some molecules will arrive at the incorrect location. It is impossible to specifically target a nanomachine while employing this hybrid communication strategy, but the diffusion technique will only have an impact on a small number of nanomachines, most of which are probably connected to that nanogateway. Yet, the nanomachines will require a method to toss out packets that lack their address [96]. Due to the ability of nanogateways to store information, such as a translation table or a routing table that will be formed, the routing idea is taken into consideration in the aforementioned nanonetwork. The proposed mechanism for the formation of a routing table takes into account the fact that each nanogateway has a unique neutral chemoattractant in addition to its own chemoattractant, which enables it to broadcast throughout the environment (similar to a broadcast address in computer networks). When the nanonetwork is put into use, the nanogateways will broadcast to the environment utilizing bacteria that are sensitive to neutral chemoattractants. The address of each nanogateway corresponds to a particular chemoattractant. Upon receiving that information, the nearby nanogateways will update their routing tables and broadcast the details of their routing tables. As a result, the nearby nanogateways will obtain the routing table, update theirs, and broadcast it again. This process will continue until a neighbor sends a nanogateway a routing table and the nanogateway realizes that no new information was provided; therefore, it does not need to update its routing table, and the neighbor does not broadcast his routing table [96]. A nanogateway will read the destination nanogateway when it receives a message and process it. If the address is its own, the message is diffused downward to its associated nanomachines; if not, the nanogateway will check its routing table, look for that destination address, and determine which gateway it needs to transfer the message to. It will query the translation table to identify what kind of bacteria to use, encode the message in the bacteria, and then send the bacteria once it has determined where the message needs to be conveyed to next. While using a diffusion process, it is not possible to precisely target a particular nanomachine. This process continues until the message reaches the destination gateway, which then sends the message to the related nanomachines. The nanomachines will therefore read their address when they receive a message; if it does not belong to them, the message is instantly destroyed; otherwise, it is read [96].

42

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

In [24], a trustworthy end-to-end MC with source, intermediate, and destination bionanomachines exchanging structured information molecules is presented. This design’s key characteristic is the ability of an intermediary bio-nanomachine to do packet replication in order to create duplicates of a molecular packet. As a result, a source bio-nanomachine creates a molecular packet and sends it into the environment. An intermediate bio-nanomachine then copies the molecular packet after detecting it. Finally, a destination bio-nanomachine creates an acknowledgment molecular packet after receiving a molecular packet and sends it back to the source bio-nanomachine. A molecular packet is a type of organized data molecule made up of header and data components. A molecular packet’s header section includes control data such as the sequence number, source address, destination address, error handling code (like the Hamming code), and others. A molecular message is transmitted by the data component between the source and destination biological machinery (or application layer processes on the source and destination bio-nanomachines). An ACK (Acknowledgement) molecular packet’s data portion is empty. DNA and RNA molecules utilized in the design of MC systems and wet laboratory implementation of MC systems may be used to implement structured information molecules. Broadcasting is a workable method to address a target more or less in conventional digital networks. Potential receivers must determine whether they are interested in the data before selecting a receiver. It can be applied to numerous novel forms of communication. For instance, calcium ions diffuse from a cell in all possible directions relying on a random walk, known as Brownian motion [82, 97], which is used in [98] employing calcium signaling in MC by passive transportation to convey data. Particles involved in collisions with atoms and molecules can travel in seemingly random, shifting motion. Data therefore spreads throughout the network like a broadcast [17]. The scalability and processing of the molecular network are also addressed in [27]. Similar to sensor networks, MC networks can have a large number of nearby bio-nanomachines that can sense the same data. Information processing and combining methods (also known as “in-network processing”) are anticipated to enhance the performance and scalability of MC networks. A slime mold, for example, has numerous sensors all over its body that allow it to assess its surroundings and determine whether they are beneficial. A method for many scattered bio-nanomachines to agree on the average value of detected concentrations is described in MC in [27, 99]. The performance and scalability of MC networks will be improved by developing signal processing techniques based on biological systems or sensor networks, as the existing MC techniques are currently somewhat limited. Diffusion channel-based MC experiences significant losses, slow propagation, and short communication ranges as a result of the unpredictability of diffusion. Chemotaxis is used to introduce directional MC in [100]. Because it keeps molecules traveling along a preset path, directional MC can dramatically increase MC efficiency. It is very appropriate in the target tracking scenario associated with various applications, such as drug delivery [41, 55, 63, 73, 89, 101–105].

2.1 Molecular Communication Network Architecture

43

A brand-new massive beacon coordinates system model is suggested in [100] to help with target tracking. With this system, beacons guide nanomachines, and the beacon system can pinpoint their precise location. To exchange information, each nanomachine transports several bacteria carriers (E. coli). Bacterial carriers transmit information encoded in DNA molecules to other nanomachines. Nanomachines can share their present location with others to enable collaborative quick target tracking with the aid of bacteria carriers. The cited research made use of a specific DNA data packet. In this scenario, DNA serves as a carrier of information encoded by nanotechnology. Like a letter in an envelope, the nanomachines package information in bacteria carriers by encoding it in DNA. In order for the nanomachines to connect with one another, the bacteria carriers act as the courier, transporting DNA from one to the other. The plasmid, a circular DNA sequence with a length of pairs ranging from 5000 to 400,000, is where the information is encoded. For ease of use, the DNA information packet format’s design shows its contents. The following details are serially stored in the DNA data packets [100]: 1. 2. 3. 4. 5.

Header. Flag that indicates the presence of target. Current coordinates of nanomachines in the beacon coordinates system. Current concentrations attractants of beacons respectively. Estimated distance to the target.

Included in the header are the host nanomachine tag, number, ACK, and check bit. The flag indicates whether the host nanomachine discovers the target. The current location of the host nanomachine is revealed by the combination of coordinates and concentrations. The nanomachine that gets the information benefits from an estimated distance to the target to assess the DNA packet’s worth more quickly [100]. A methodology for opportunistic routing in diffusion-based MC (OR-DMC) based on concentration gradient and distance information has been put out in [106]. A straightforward multi-hop communication technique-based diffusionbased molecular nanonetwork with a number of nanonodes that connect with a nanogateway has been considered. For information sharing, a pulse-based modulation method is employed. A nanonode can transfer information by sending out a pulse of molecules that will cause the transmission medium to spike. In ORDMC, two methods of pulse-based modulation have been applied. Energy detection comes first; then the receiver nanonode will measure it. The observed energy will be measured as a specific time or a molecule concentration integral. The received pulse energy will be assessed against the predetermined threshold. Amplitude detection is the second method. The recipient nanonode tracks changes in the local molecule concentration over a predetermined period of time. The received signal is then processed and put in a threshold value comparison with the maximum concentration. Pulse amplitude is the term for this highest concentration [106]. In contrast to existing single-hop routing protocols, OR-DMC selects the highest priority forwarder by contrasting all potential forwarders utilizing various infor-

44

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

mation exchange kinds. Moreover, a coordination method is used to guarantee the specific forwarder. The OR-DMC uses distance information and thankfulness for concentration in its two phases. Every nanonode determines its distance from every other nanonode that is within its communication range throughout the training phase of the process. The messaging is sent by nanonodes during the routing phase, which comes next. In OR-DMC, it has been assumed that every nanonode in the communication range is a member of the forwarder set. This makes the process of creating a forwarder set simple and obstacle-free. As indicated in Fig. 2.10, a twohop situation is taken into consideration to exchange information from the sender to the gateway, which is outside of its communication range [106]. It is always considered that the nanonode closest to the sender node is the optimal node to choose as a next hop when choosing the next forwarder. Just distancebased routing, however, won’t ensure that the message reaches the gateway. The concentration gradient has been used as another parameter to guarantee this. Beacon nodes, a different kind of nanonode that OR-DMC presume is more sophisticated than conventional nanonodes, are used. The beacon node is a gateway node that periodically releases high concentrations of beacons into the environment that may be seen by every nanonode in the nanonetwork during the beaconing period. Information on the gradient in concentration is also obtained. The length of the beaconing period is determined by the size of the intended use; the nanonode sends

Fig. 2.10 Proposed model of OR-DMC, ACK, timing sequence, and different types of messages are shown

2.1 Molecular Communication Network Architecture

45

a concentration signal to the nanonetwork, which is followed by transmission slots with waiting periods before transmitting a molecular spike to measure distance. Each nanonode continuously monitors the concentration. This expanded approach will track and offer distance data similar to beacons. Each message includes orientation and focus information. This will conclude with the signal. This is being sent out by a sender node. Here, the nanonode will not only be aware of the information regarding the distance but also the concentration signal will decide the nanonodes closer to the will also be aware of the quantity of nanonodes in its forwarder set, gateway node within the communication range, and takes part which are situated closer to the gateway in the message forwarding. A straightforward acknowledgment-based coordination technique has been employed to confirm that the message has been forwarded. An ACK is generated by the highest priority node if it successfully transmits the message packet; otherwise, no ACK is generated if the message packet is not successfully transmitted within a predetermined amount of time. The message will be transmitted by the nanonode with the next higher priority. The intermediate nanonodes in the proposed OR-DMC must decide the following [106]: a. Eligibility to take part in the forwarding process or not. This eligibility depends upon the concentration signal. b. Priority to take part in the forwarding process. This priority will be dependent on time slots to wait. The training step is improved by OR-DMC in order to assess the priority of eligible nanonodes. Information on the concentration gradient is also collected rather than distance. Before transmitting the chemical spike for determining distance, the nanonode delivers a concentration signal for the aforementioned reason. Both distance and orientation information will be provided by this enhanced method. By the end of this, the nanonode will be aware of the number of nanonodes in its forwarder set that are situated closer to the gateway in addition to the distance information [106].

2.1.4 Molecular Link Layer Functionalities for communication within a direct range of communication are provided by the molecular link layer (e.g., a direct range of signal propagation) [7]. A group of bio-nanomachines can successfully transmit information with one another within the confines of a communication range using the mechanisms offered by the link layer of an MC architecture. An environment where information molecules can travel from a transmitter bio-nanomachine to a receiver bionanomachine in a reasonably short length of time to trigger biochemical reactions at the receiver bio-nanomachine is basically what is meant by a communication range. The link layer offers mechanisms for handling errors, regulating transmission

46

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

speed (flow control), sharing a medium (media access control), addressing recipient bionanomachines, synchronizing clocks in bionanomachines, and determining the distance between bionanomachines [8].

2.1.4.1

Molecular Link Layer Communication Model

The signaling sublayer is followed logically by the molecular link layer, which interacts with the physical materials (Fig. 2.2). (i.e., the physical layer). The molecular link layer’s components are all rational components. The following are the fundamental elements of the molecular link layer, which are extremely similar to those of the signaling sublayer [7]: 1. Molecular frame. It is an idea, fact, piece of information, or statement that both the transmitter and the receiver comprehend at the molecular link layer. It is a molecular connection layer that functions as the signaling sublayer’s message [7]. 2. Sender. A molecular frame is sent. With the exception of the fact that a sender at the molecular link layer does not perform signal modulation, it is a molecular link layer equivalent to the transmitter at the signaling sublayer [7]. 3. Receiver. A sender sends a molecular frame, which it receives. A molecular link layer receiver is comparable to the receiver at the signaling sublayer, with the exception that it does not perform signal demodulation [7]. 4. Molecular communication link. It is the logical channel through which a molecular frame is transmitted between a sender and a receiver who are in direct communication. It is a molecular link layer that functions as the signaling sublayer’s MC channel. Several MC channels linking the same sender/receiver pair at the signaling sublayer may make up a single MC link between a sender and a receiver. Point-to-point links and shared medium links are the two different forms of MC links, similar to the MC channel at the signaling sublayer. Many senders and numerous receivers are connected by a shared medium link. For instance, a broadcast MC channel at the physical layer may be used to implement a shared medium link at the molecular link layer (the signaling sublayer). A single sender and a single receiver are linked together by a point-to-point link. For example, a point-to-point MC channel at the physical layer might be used to build a point-to-point connection at the molecular link layer (the signaling sublayer) [7]. 5. Molecular frame storage. Keeping molecular frames at a transmitter or receiver while that device processes another molecular frame is a logical part of the process. Using either molecule storage at the bio-nanomachine sublayer or signal molecule storage at the signaling sublayer, molecular frame storage can be accomplished. When molecular frames in the molecular frame storage decay and lose their functionality over time or when the molecular frame storage reaches its maximum, there is a loss of molecular frames [7].

2.1 Molecular Communication Network Architecture

2.1.4.2

47

Molecular Link Layer Functionalities

1. Framing. It is the capability to convert signal molecules from the physical layer into a molecular frame. A molecular frame that framing creates also receives a unique identification. The functions of the physical layer may allow framing at the molecular link layer [7]: • Vesicle-based framing. A vesicle may naturally correlate to a molecular frame at the molecular link layer when the bio-nanomachine sublayer supports the functionality of “release molecules” by releasing vesicles containing numerous types of molecules and functionality of “catch molecules” by merging with vesicles. Signal molecules that serve as messages may be found in vesicles, along with other molecules that serve as headers for the molecular link layer. A molecule acting as an identifier could be used to tag a vesicle [7]. • DNA-based framing. A DNA molecule that has a message and an errorcorrecting code in its sequence (i.e., molecular link layer information) may correspond to a molecular frame at the molecular link layer if the bionanomachine sublayer supports functionalities of “release molecules” and “capture molecules” of DNA molecules. A DNA molecule may also include a molecular frame identifier as part of its sequence [7]. 2. Addressing. The capability to designate a receiver (or a group of receivers) that receives molecular frames or a location (or a number of locations) to which molecular frames are sent is present at the molecular link layer. Addressing at the molecular link layer can support a physical address (i.e., an address that identifies a receiver or a group of receivers) and a location address, similar to addressing at the signaling sublayer (i.e., an address that identifies a location or multiple locations). It may also support a group address as well as an individual address (i.e., an address that designates a single recipient or a single place) (i.e., an address that identifies a group of receivers or multiple locations) [7]. Logical addresses at the molecular link layer map to addresses at the physical layer (i.e., physical addresses and location addresses at the signaling sublayer, in Fig. 2.2). The physical layer receives information from the molecular link layer regarding the address, the type of address it employs (physical address, location address, individual address, group address), and group membership (if it is a group address) (i.e., the signaling sublayer) [7]. Moreover, the link layer offers addressing methods that let a sender biological machine designate specific receiver biological machines inside of a communication range. In order to communicate with the reception bio-nanomachines, a sender bio-nanomachine may employ the addresses connected to those machines. These addresses can be implemented using biological components like a pair of complementary DNA sequences. If one DNA sequence is attached to an information molecule and the other sequence is attached to the receiver bio-nanomachine, the information molecule binds to the receiver bio-nanomachine through the pairing of DNA sequences. When communicating with bio-nanomachines at a particular spot within a communication range, a sender machine may also

48

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

employ location addresses. In developmental biology, a place is addressed by concentrations of molecules there, and a biological cell’s development advances in accordance with the location. Similar to MC, information molecules may be built to propagate toward a destination indicated by a set of concentrations of molecules, and a location in a communication range may be addressed by a set of concentrations of molecules [8, 25]. 3. Molecular frame transmission/reception. It is a capability to send (receive) a molecular frame across an MC link to its recipient (from its sender). Its functionality is dependent on the MC link type being used [7]: • Point-to-point molecular communication link. Physical layer functionalities like “signal molecule transmission/reception” (to transmit/receive a molecular frame), “signal propagation” (to propagate a molecular frame over a point-to-point link), and “signal relay” may support the relatively simple process of sending and receiving molecular frames over point-to-point links (to propagate a molecular frame to its receiver over long distance). The physical layer receives from the molecular link layer the address of the receiver for a molecular frame [7]. • Shared medium molecular communication link. Media access control is necessary for sending and receiving molecular frames across a shared link (MAC). A shared MC link is divided among several senders via media access control, and molecular frames are transmitted from various senders onto a shared link without interfering with one another. Time-division multiplexing (TDM), which allows distinct senders to transmit molecular frames at various times, is one method of medium access control. Another method of limiting access to a medium is carrier sensing, in which a transmitter first checks its surroundings for other molecular frames before transmitting a molecular frame [7]. Physical layer features may assist molecular link layer media access control. For instance, bio-nanomachine sublayer functions of “keep track of time” and “release molecules” in a controller manner may provide TDMbased media access control. Bio-nanomachine sublayer functionalities such as “capture molecules” to detect the presence of a certain type of molecules (i.e., a molecular frame) in the environment, as well as “keep track of time” and “release molecules” in a controlled manner, may support carrier sense-based media access control. Information on when to release molecules is transmitted to the bio-nanomachine sublayer by the molecular link layer [7]. 4. Molecular frame loss handling. Molecular link layer loss of a molecular frame is handled by this functionality. When molecular frames in the molecular frame storage deteriorate and lose their functionality over time or when the molecular frame storage reaches its limit, the loss may happen at the receiver and affect the molecular link layer. The two tasks involved in molecular frame loss management are loss detection and loss correction via retransmission. Each task may be sender-initiated or receiver-initiated depending on who starts it. Below is a description of each task [7]:

2.1 Molecular Communication Network Architecture

49

• Molecular frame loss detection. A receiver may notice the loss of a molecular frame if it happens there (receiver-initiated loss detection). The following method can also be used by the sender to trigger loss detection. For instance, when a molecular frame is lost at a receiver, the receiver may either fail to produce the predicted biochemical events or produce them to a lesser extent than expected. The sender may detect the existence or absence of anticipated biological reactions within a timeout period as a result of biochemical reactions the receiver creates that may spread via the environment to the sender. If the molecular frames the sender communicated successfully reach the receiver, it may be possible to tell by the strength of the anticipated biological responses in the surroundings. Physical layer features like “detect molecules” may assist sender-initiated loss detection. The physical layer may receive information from the molecular link layer regarding biochemical conditions to be detected [7]. • Molecular frame retransmission. Retransmission may begin when a sender notices the loss of a molecular frame (sender-initiated retransmission). Physical layer features such as “modify molecules” and “release molecules” may facilitate sender-initiated retransmission. The information on molecular frames may be passed from the molecular link layer to the physical layer for retransmission. Retransmission may begin when a receiver notices the loss of a molecular frame (receiver-initiated retransmission). The receiver tells the sender when it discovers the loss. In order to do this, the receiver must be clever enough to recognize a loss and inform the sender [7]. 5. Molecular error handling. As information molecules are transmitted from a sender bio-nanomachine to a receiver bio-nanomachine, the molecular link layer offers means to deal with potential mistakes that may arise. Information molecules may deteriorate in the environment or arrive with significant jitter (i.e., dispersed arrival times and in a different order from the original transmission), which could cause unexpected receiver reactions, or errors. Error handling may be necessary. A sender biological machine may send an excessive number of information molecules, just like in biological systems. Only after detecting a predetermined quantity of these chemicals does the receiver biological machine respond. The influence of noise or fluctuation is diminished when there are more information molecules present because the signal-to-noise ratio rises. To prevent information molecules from degrading during propagation and reduce error rates, an interface molecule may also enclose information molecules. Vesicles, which are employed as a container to transport things inside a biological cell, could be used to implement such an interface molecule. A pattern of molecules conveying information may also allow for the encoding of error detection and correction codes. For instance, Hamming codes could be contained in the bit order of an information molecule. For error management at a receiving biological machine, it would also be conceivable to incorporate such codes into the structure of an information molecule (such as a DNA sequence) [8, 25].

50

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

6. Molecular frame flow control. It is a capability for a sender to change the pace at which molecular frames are transmitted in order to prevent loss at a receiver. For applications like targeted drug delivery, where the signal molecules (e.g., drug molecules) are expensive and where an excess amount of signal molecules can have undesirable side effects, it is crucial to prevent the sender from transmitting an excessive number of molecular frames (or signal molecules). Molecular frame flows control at the molecular link layer can be sender-initiated or receiver-initiated depending on who initiates flow control. This is similar to molecular frame loss handling [7]. • Sender-initiated flow control. The simplest type of sender-initiated flow control is when a sender transmits molecular frames at a very low rate, ensuring that no frames are lost. A prior understanding of the receiver’s capacity to process incoming molecular frames is necessary for this. Such knowledge may be used to create bio-nanomachines and the MC layered architecture, or external control may supply such knowledge with the molecular link layer [7]. Sender-initiated flow control might be more intricate when the sender notices the loss of a molecular frame. The environment may be sensed by a sender to determine whether or not the biochemical reactions that a receiver is predicted to produce are present. The sender then modifies the rate of molecular frame transmission in accordance with the predicted strength of the biochemical reactions in the environment. Physical layer features like “modify molecules,” “detect molecules,” and “release molecules” may provide senderinitiated flow control (in a controlled manner to adjust the rate of molecular frame transmission) [7]. The physical layer may receive information from the molecular link layer regarding the rate of transmission of molecular frames [7]. • Receiver-initiated flow control. The receiver may alert the sender when the molecular frame storage at a receiver begins to fill up, and the sender may then modify its transmission rate. The link layer of conventional layered communication networks is where this type of flow control is most frequently used. In order to determine the amount of molecular frame storage space that is available and to provide feedback to the sender, the receiver must be intelligent [7]. 7. Molecular frame synchronization. In bio-nanomachines, the molecular connection layer also offers clock synchronization mechanisms. Biochemical clocks, such as circadian clocks, are used by biological systems, and they may also be included in bio-nanomachines. Biochemical clocks can be artificially inserted into bacterial cells, as synthetic biology shows. When a set of bio-nanomachines complete a task simultaneously or to prevent interference during time-slotted communication, clock synchronization may be advantageous. It is possible to synchronize the clocks of biological nanomachines because biological systems have examples of clock synchronization such as heart cells contracting at the

2.1 Molecular Communication Network Architecture

51

same time, bacteria using quorum sensing to determine when to form a film, and sequential cell differentiation during developmental growth [8, 25]. 8. Molecular distance measurement. Mechanisms for determining the distance to bio-nanomachines in the environment are also provided by the molecular link layer. The distance data may be helpful for adjusting the spread of bionanomachines, pinpointing their relative locations, or maximizing transmission rates (e.g., for flow control). In electronic radio networks, a pair of transceivers that communicate using radio waves can determine how far apart they are from one another by measuring the time-of-flight and signal attenuation of the radio waves. Similar methods may be used in MC since the concentration of molecules drops with distance and the predicted time for a molecule to propagate grows with distance (i.e., a property comparable to time-of-flight) (i.e., a characteristic similar to signal attenuation) [8, 25]. 9. Storing molecular frames. The corresponding molecular frame storage of the sender and the receiver may be used to store molecules [7].

2.1.4.3

Significant Cases of Investigations in Molecular Link Layer

The research of a MAC (medium access control) protocol and its impact on the interaction of many molecular transmitters in the same environment is presented in [107]. Because of the physical interactions that can take place between molecules released by various transmitters, such as collisions or electrical and chemical reactions, the performance of the MC system in terms of attenuation and delay will probably vary. In order to reduce interference between various emitters and increase the network’s overall capacity, a MAC protocol will be needed. Also, in order to facilitate communication between various sources and destination points, routing and addressing requirements must be met. In our vision, any type of addressing, such as their type or electrical charge, will probably be encoded inside the structure of the molecules that make up the information message. A communication model between two bacterial populations is taken into account in [30]. A group of bacteria encased in a chamber is referred to in this paradigm as a node. The communication channel, transmitter node, and receiver node make up the model. It is believed that both the transmitter and receiver nodes are genetically altered microorganisms. Three processes can be used to communicate molecules between two bionodes. Thus, the receiver node detects the concentration of the local signaling molecules and responds appropriately once the transmitter node creates the signaling molecules by proper stimulation, after which the molecules travel across the medium while following Brownian motion. Diffusion is necessary for the transmission of the information since it is embedded by altering the concentration of the signaling molecules, which is how the information is encoded during transmission. To determine the concentration of signaling molecules nearby the receiver node and subsequently decode the transmitted data, the output of the receiver node is measured in steady state as luminescence [30].

52

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

In the previously cited work [30], to effectively represent the transmitted symbols, the propagation time is divided into time slots (also called symbol duration), which have equal lengths. Only one symbol propagates in a single time slot. The communication channel is a binomial one, where each molecule arrives at the receiver or does not. As in typical communication channels, undesired effects affect the information; therefore, the previous bits can influence the current bit, produce intersymbol interference (ISI), and generate BER (bit error rate) at the destination. Frequent packet corruptions and out-of-sequence delivery are caused by a timevarying channel with a relatively high BER level, which calls for error check codes and Automatic Repeat reQuest (ARQ) methods for effective error detection and recovery, respectively. The data link layer, which receives packets from the network layer and wraps them into frames for transmission, typically performs the functions of an error-correcting ARQ [17]. Because of this, each molecular frame in [30] includes a frame header, a payload field for the packet, and a frame trailer like the one in the top panel of Fig. 2.11. Peer-to-peer protocols that offer dependable

Fig. 2.11 Structure of molecular frames and techniques for error in bacterial quorum communications

2.1 Molecular Communication Network Architecture

53

information transport are built on the ARQ protocol foundation. The Stop-N-Wait (SW-ARQ) method, the simplest ARQ scheme, is utilized in the communications paradigm to enhance channel performance. A frame is sent to the receiver in SWARQ, and the transmitter node then waits for the receiver to acknowledge the frame. A series of information frames are created by the transmitter node for transmission. The header of the frame must contain control information necessary for the ARQ mechanism to function properly. Moreover, CRC check bits will be added to the frame to check for transmission errors. This is done because CRC (cyclic redundancy check) has a good error-sensing capability as well as quick encoding and decoding capabilities. The fundamental components of the ARQ mechanism are shown in the lower panel of Fig. 2.11. It includes the acknowledgment frames (ACKs), the time-out mechanisms, and the information frames that convey the data packets. The ACK frame indicates that a specific frame has been received. The time-out mechanism is necessary to keep the frames moving. This model makes the assumption that data only transmits in one direction, from transmitter to receiver. The transmission of ACKs is the only use of the reverse communication channel. The authors of [108] describe .μ-NET, a microfluidic local area network (LAN) that enables the exchange of biochemical and digital data conveyed by droplets traveling across molecular processors (droplet microfluidic devices) in a microfluidic chip. DNA, RNA, and protein biosynthesis are examples of molecular biology applications that can be supported by the -NET. The biological LAN incorporates methods for medium access control, switching, and addressing information. In fact, the distance between droplets in .μ-NET is associated with the address of the molecular processor to which a droplet should be transported. Switching is used to direct the droplets inside the microfluidic device, and medium access control is used to prevent droplet collisions that could cause them to fuse and lose their biological information. As shown in Fig. 2.12, the data passing between the network

Fig. 2.12 Microfluidic frame carried by continuous phase in a microfluidic channel

54

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

nodes in this biological LAN is structured into microfluidic frames made up of two droplets: a header droplet and a payload droplet. The header droplet is used for signaling, whereas the payload droplet contains the bio/chemical data that will be processed by the molecular processor. In fact, the address of the molecular processor that must receive the payload droplet is digital information encoded in the distance between the header and payload droplets [108]. The header droplet is used to regulate the proper routing of the payload to the network element. The authors describe a strategy known as droplet sense multiple access (DSMA), which is similar to the carrier sense multiple access (CSMA) employed in conventional communication networks, in [108]. The frame structure used in conventional networks, which consists of a header, correct information, and a trailer [62], served as inspiration for the molecule scheme in [109], which introduces a method for encoding a complete data frame in a messenger molecule’s chemical structure. This method is known as a Molecule-asa-Frame (MaaF). An overhead with frame identity bits and an information-carrying payload is contained within the frame. The frame identification bit combinations serve as sequence numbers for succeeding frame broadcasts once the receiver and transmitter have synchronized. In [109], it primarily takes into account encoding bits into the interior chemical structures of messenger molecules. The transmitter may synthesize molecules matching the bit string to be conveyed, or it may use other qualities like optical isomeries found in a family of molecules, to encode multiple bits. In order to generalize this claim, MaaF suggests equating a family of molecules with multiple bit-carrying capabilities to a communication frame with multiple accessible bits. A bit slot in a frame is a component that is both changeable and recognizable by the receiver and transmitter. Figure 2.13 shows a molecule family with n available bit slots. It is important to note that the frame-based MaaF technique can be used to any family of molecules with a set number of available bit slots for encoding. In a hypothetical frame, any collection of numerous distinct molecules can be thought of as a variety of bit string combinations. This suggests that MaaF is a general strategy that may be expanded for any situation involving various molecule types, making

Fig. 2.13 A general structure of a MaaF frame with n bit slots

2.1 Molecular Communication Network Architecture

55

Fig. 2.13 applicable in all such situations. Figure 2.13 illustrates the relationship between n-squares and n-bit slots. These are the frame slots that the transmitter can modify. And observe that the black shapes are distinct on either end. This allows the receiver to distinguish between the bit string’s head and tail [109]. The MAC-Address notion (i.e., a hardware framework utilized for local addressing) is mentioned in [98]. It is used in a biological investigation and typically consists of 48 manufacturer and individual unique address bits. For a nanomachine, an equivalent with hardware-encoded address data is necessary. In a biological sense, it might also be adapted by particular cell data, like distinct DNA strands. The messenger molecule could use DNA scaffolding to selectively attach at the required site by using a complementary DNA strand. The possibility of using self-assembly processes to assemble nanoparticles and nanomachines into bigger structures is currently being thoroughly examined. The reuse of these strands requires helicase enzyme activity to separate the complementary strands losslessly after message delivery for the next reception since it can give unique addresses for nanomachines. The authors in [26] take into account an MC between two nanoscale devices with internal clocks but no external clock signal. Hence, notwithstanding their misalignment, symbols are broadcast in accordance with the transmitter clock and detected in accordance with the receiver clock. In order to start the communication, the transmitter specifically delivers a particular chemical signal called a beacon to start the receiver’s detecting process. In such cases, even with the misalignment of the clocks, this technique does not result in a lot of errors if the sequence of symbols is short. The well-known asynchronous serial communication utilized in traditional networks has a molecular equivalent in this approach. The start bit (framing symbol) specifically serves the role of the beacon sign. In addition to this scheme’s inherent simplicity, the authors of [26] believe that it meets the needs of many applications that are envisioned, particularly those in the field of nanomedicine, which are likely to rely on the exchange of brief but efficient messages between intra-body nanomachines [110]. The suggested molecular asynchronous serial communication presupposes that both the transmitter and the receiver clocks are flawless, though not synced, according to additional considerations in [26]. As a result, the study may concentrate solely on how beacon detection mistakes affect whether future information symbols are correctly detected (via the symbol error probability). These framing mistakes are inevitable due to the beacon arrival instant’s randomness, which typically results in an unfavorable shift between the information symbol’s real and ideal detection instants. The aforementioned research takes into account a multilayer molecular encoding method, where each symbol is represented by a chemical substance, for the purpose of generality. Thus, in [26], the authors note that, also assuming a realistic molecular channel with any amount of ISI, the complexity of implementation will ultimately dictate plausible values for the number of symbols. They then arrive at a performance function based on these presumptions that convert beacon synchronization faults (frame errors) into information symbol errors (via the symbol error probability). Lastly, they established standards for producing a beacon symbol with an extremely

56

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

low error probability. The emission intensity and diffusion coefficient of the beacon molecules must be chosen for this design.

References 1. M. Dong, W. Li, X. Xu, Evaluation and modeling of hiv based on communication theory in biological systems. Infect. Genet. Evol. 46, 241–247 (2016) 2. N. Farsad, Y. Murin, A. Eckford, A. Goldsmith, On the capacity of diffusion-based molecular timing channels, in 2016 IEEE International Symposium on Information Theory (ISIT) (IEEE, 2016), pp. 1023–1027 3. N.A. Ali, M. Abu-Elkheir, Internet of nano-things healthcare applications: Requirements, opportunities, and challenges, in 2015 IEEE 11th International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob) (IEEE, 2015), pp. 9–14 4. A. Dubey, S. Tandon, A. Seth, Design of a molecular communication framework for nanomachines, in 2012 Fourth International Conference on Communication Systems and Networks (COMSNETS) (IEEE, 2012), pp. 1–2 5. J.-T. Huang, H.-Y. Lai, Y.-C. Lee, C.-H. Lee, P.-C. Yeh, Distance estimation in concentrationbased molecular communications, in 2013 IEEE Global Communications Conference (GLOBECOM) (IEEE, 2013), pp. 2587–2591 6. Q.H. Abbasi, K. Yang, N. Chopra, J.M. Jornet, N.A. Abuali, K.A. Qaraqe, A. Alomainy, Nano-communication for biomedical applications: A review on the state-of-the-art from physical layers to novel networking concepts. IEEE Access 4, 3920–3935 (2016) 7. T. Nakano, T. Suda, Y. Okaie, M.J. Moore, A.V. Vasilakos, Molecular communication among biological nanomachines: A layered architecture and research issues. IEEE Trans. Nanobiosci. 13(3), 169–197 (2014) 8. T. Nakano, A.W. Eckford, T. Haraguchi, Molecular Communication (Cambridge University Press, Cambridge, 2013) 9. A.Z. Shirazi, S.M. Mazinani, S.K. Eghbal, Protocol stack for nano networks, in 2012 International Symposium on Computer, Consumer and Control (IEEE, 2012), pp. 849–853 10. B.J. Wysocki, T.M. Martin, T.A. Wysocki, A.K. Pannier, Modeling nonviral gene delivery as a macro-to-nano communication system. Nano Commun. Netw. 4(1), 14–22 (2013) 11. U. Chude-Okonkwo, R. Malekian, B. Maharaj, Internet of things for advanced targeted nanomedical applications, in Advanced Targeted Nanomedicine (Springer, 2019), pp. 113– 124 12. F. Walsh, S. Balasubramaniam, D. Botvich, T. Suda, T. Nakano, S.F. Bush, M. Foghlú, Hybrid DNA and enzyme based computing for address encoding, link switching and error correction in molecular communication, in International Conference on Nano-Networks (Springer, 2008), pp. 28–38 13. A. El-taweel, S.M. Abd El-atty, S. El-Rabaie, Efficient molecular communication protocol based on mobile ad-hoc nanonetwork. Menoufia J. Electron. Eng. Res. 26(2), 427–443 (2017) 14. D. Bi, A. Almpanis, A. Noel, Y. Deng, R. Schober, A survey of molecular communication in cell biology: Establishing a new hierarchy for interdisciplinary applications. IEEE Commun. Surv. Tutorials 23(3), 1494–1545 (2021) 15. Y. Cevallos, L. Molina, A. Santillán, F. De Rango, A. Rushdi, J.B. Alonso, A digital communication analysis of gene expression of proteins in biological systems: A layered network model view. Cogn. Comput. 9(1), 43–67 (2017) 16. S. Balasubramaniam, S. Ben-Yehuda, S. Pautot, A. Jesorka, Y. Koucheryavy et al., A review of experimental opportunities for molecular communication. Nano Commun. Netw. 4(2), 43– 52 (2013)

References

57

17. O. Tsave, I. Kavakiotis, K. Kantelis, S. Mavridopoulos, P. Nicopolitidis, G. Papadimitriou, I. Vlahavas, A. Salifoglou, The anatomy of bacteria-inspired nanonetworks: Molecular nanomachines in message dissemination. Nano Commun. Netw. 21, 100244 (2019) 18. T. Nakano, Molecular communication: A 10 year retrospective. IEEE Trans. Mol. Biol. MultiScale Commun. 3(2), 71–78 (2017) 19. A.D. Odysseos, T. Nakano, Bionanomachine diagnostics and nanonetwork therapeutic in brain malignancies with bionanodevice interfaces. IEEE Trans. Mol. Biol. Multi-Scale Commun. (2021). https://doi.org/10.1109/TMBMC.2021.3083725 20. J.T. Gómez, K. Pitke, L. Stratmann, F. Dressler, Age of information in molecular communication channels. Digit. Signal Process. 124, 103108 (2022) 21. M. Egan, M. Kuscu, M.T. Barros, M. Booth, A. Llopis-Lorente, M. Magarini, D.P. Martins, M. Schäfer, P. Stano, Toward interdisciplinary synergies in molecular communications: Perspectives from synthetic biology, nanotechnology, communications engineering and philosophy of science. Life 13(1), 208 (2023) 22. N. Farsad, H.B. Yilmaz, A. Eckford, C.-B. Chae, W. Guo, A comprehensive survey of recent advancements in molecular communication. IEEE Commun. Surv. Tutorials 18(3), 1887– 1919 (2016) 23. T. Suda, T. Nakano, Molecular communication as a biological system, in 2018 IEEE International Conference on Sensing, Communication and Networking (SECON Workshops) (IEEE, 2018), pp. 1–4 24. T. Furubayashi, Y. Sakatani, T. Nakano, A. Eckford, N. Ichihashi, Design and wet-laboratory implementation of reliable end-to-end molecular communication. Wirel. Netw. 24(5), 1809– 1819 (2018) 25. M.N. Anjum, H. Wang, Molecular Communication for Wireless Body Area Networks (2019) 26. B. Atakan, S. Galmes, Effects of framing errors on the performance of molecular communications with memory. IEEE Access 8, 19,970–19,981 (2020) 27. T. Nakano, M.J. Moore, F. Wei, A.V. Vasilakos, J. Shuai, Molecular communication and networking: Opportunities and challenges. IEEE Trans. Nanobiosci. 11(2), 135–148 (2012) 28. I.F. Akyildiz, F. Fekri, R. Sivakumar, C.R. Forest, B.K. Hammer, Monaco: fundamentals of molecular nano-communication networks. IEEE Wirel. Commun. 19(5), 12–18 (2012) 29. N. Dalchau, G. Szép, R. Hernansaiz-Ballesteros, C.P. Barnes, L. Cardelli, A. Phillips, A. Csikász-Nagy, Computing with biological switches and clocks. Nat. Comput. 17(4), 761– 779 (2018) 30. C. Bai, M.S. Leeson, M.D. Higgins, Performance of SW-ARQ in bacterial quorum communications. Nano Commun. Netw. 6(1), 3–14 (2015) 31. Y. Cevallos, L. Tello-Oquendo, D. Inca, N. Samaniego, I. Santillán, A.Z. Shirazi, G.A. Gomez, On the efficient digital code representation in DNA-based data storage, in Proceedings of the 7th ACM International Conference on Nanoscale Computing and Communication (2020), pp. 1–7 32. Y. Cevallos, L. Tello-Oquendo, D. Inca, C. Palacios, L. Rentería, Genetic expression in biological systems: A digital communication perspective. Open Bioinf. J. 12(1), 45–49 (2019) 33. Y. Cevallos, T. Nakano, L. Tello-Oquendo, A. Rushdi, D. Inca, I. Santillán, A.Z. Shirazi, N. Samaniego, A brief review on DNA storage, compression, and digitalization. Nano Commun. Netw. 31, 100391 (2022) 34. Y. Cevallos, T. Nakano, L. Tello-Oquendo, D. Inca, I. Santillán, A.Z. Shirazi, A. Rushdi, N. Samaniego, Modeling gene expression and protein delivery as an end-to-end digital communication system. Open Bioinf. J. 14(1), (2021) 35. Y. Cevallos, T. Nakano, L. Tello-Oquendo, N. Chopra, A.Z. Shirazi, D. Inca, and I. Santillán, Theoretical basis for gene expression modeling based on the IEEE 1906.1 standard, in International Conference on Bio-inspired Information and Communication Technologies (Springer, 2021), pp. 145–162

58

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

36. T. Nakano, M. Moore, A. Enomoto, T. Suda, Molecular communication technology as a biological ICT, in Biological Functions for Information and Communication Technologies (Springer, 2011), pp. 49–86 37. L. Khaloopour, M. Mirmohseni, M. Nasiri-Kenari, Joint sensing, communication and localization of a silent abnormality using molecular diffusion. Preprint (2022). arXiv:2203.16641 38. Z. Lu, Towards a progressive e-health application framework, Ph.D. Dissertation, King Abdullah University of Science and Technology, 2022 39. I. Isik, M.B. Er, E. Isik, Analysis and classification of the mobile molecular communication systems with deep learning. J. Ambient Intell. Humaniz. Comput. 13(5), 2903–2919 (2022) 40. L. Chouhan, P.K. Upadhyay, Effect of time-dependent drift and diffusion in molecular communication systems. IEEE Commun. Lett. 26(4), 778–782 (2022) 41. L.F. Borges, M.T. Barros, M. Nogueira, Toward reliable intra-body molecular communication: An error control perspective. IEEE Commun. Mag. 59(5), 114–120 (2021) 42. S.S. Somathilaka, D.P. Martins, S. Balasubramaniam, Information flow of cascading bacterial molecular communication systems with cooperative amplification, in IEE ICC 2022 (IEEE, 2022) 43. Y. Sun, H. Bian, Y. Chen, A photolysis-assist molecular communication for tumor biosensing. Sensors 22(7), 2495 (2022) 44. D. Jing, Y. Li, A.W. Eckford, An extended Kalman filter for distance estimation and power control in mobile molecular communication. Preprint (2022). arXiv:2205.09596 45. A. Awad, M.M. Fouda, M.M. Khashaba, E.R. Mohamed, K.M. Hosny, Utilization of mobile edge computing on the internet of medical things: A survey. ICT Express (2022) 46. O. Yetimoglu, M.K. Avci, B.C. Akdeniz, H.B. Yilmaz, A.E. Pusane, T. Tugcu, Multiple transmitter localization via single receiver in 3-d molecular communication via diffusion. Digit. Signal Process. 124, 103185 (2022) 47. M. Femminella, G. Reali, Implementation issues of diffusion-based molecular communications receivers based on transcriptional elements, Digit. Signal Process. 124, 103160 (2022) 48. Y. Bayraktar, E. Isik, I. Isik, A. Ozyilmaz, M. Toprak, F. Kahraman Guloglu, S. Aydin, Analyzing of Alzheimer’s disease based on biomedical and socio-economic approach using molecular communication, artificial neural network, and random forest models. Sustainability 14(13), 7901 (2022) 49. C. Fonseca, M.T. Barros, A. Odysseos, S. Kidambi, S. Balasubramaniam, Quasi-Spherical Absorbing-Reactive Receiver Model of Glioblastoma Cells for Exosome-Based Molecular Communications 50. D. Vimalajeewa, S. Balasubramaniam, D.P. Berry, G. Barry, Virus particle propagation and infectivity along the respiratory tract and a case study for SARS-CoV-2. Sci. Rep. 12(1), 1–14 (2022) 51. A.S. Tanenbaum, D. Wetherall, The TCP/IP reference model, pp. 31, 43, 45, 46, 194, 216 (2012) 52. A. Zadeh Shirazi, E. Fornaciari, M.D. McDonnell, M. Yaghoobi, Y. Cevallos, L. TelloOquendo, D. Inca, G.A. Gomez, The application of deep convolutional neural networks to brain cancer images: a survey. J. Pers. Med. 10(4), 224 (2020) 53. T. Nakano, Y. Okaie, T. Hara, Applications of Molecular Communication Systems (Springer International Publishing, Cham, 2020), pp. 32–37 54. Z. Cheng, J. Yan, J. Sun, Y. Tu, K. Chi, Joint optimizations of relays locations and decision threshold for multi-hop diffusive mobile molecular communication with drift. IEEE Trans. NanoBiosci. 21(3), 454–465 (2022) 55. H.K. Rudsari, M. Zoofaghari, M. Veletic, J. Bergsland, I. Balasingham, The end-to-end molecular communication model of extracellular vesicle-based drug delivery. Preprint (2022). arXiv:2207.01875 56. IEEE recommended practice for nanoscale and molecular communication framework, in IEEE Std 1906.1-2015 (2016), pp. 1–64 57. IEEE standard data model for nanoscale communication systems, in IEEE Std 1906.1.1-2020 (2020), pp. 1–142

References

59

58. A.S. Tanenbaum, D. Wetherall, et al., Computer Networks (Prentice Hall, Boston, 1996), pp. 41–49 59. M. Günes, D. Reina, J. Campos, S. Toral, Mobile Ad Hoc Network Protocols Based on Dissimilarity Metrics, ser. SpringerBriefs in Electrical and Computer Engineering (Springer International Publishing, Berlin, 2017). [Online]. Available: https://books.google.com.ec/ books?id=g6s2DwAAQBAJ 60. A. Kumar, B. Xie, Handbook of Mobile Systems Applications and Services, ser. Mobile Services and Systems (CRC Press, Boca Raton, 2016). [Online]. Available: https://books. google.com.ec/books?id=AqlRErA8-a8C 61. D. Comer, Computer Networks and Internets, ser. Always Learning. [Online]. Available: https://books.google.com.ec/books?id=PBhAngEACAAJ 62. L. Felicetti, M. Femminella, G. Reali, T. Nakano, A.V. Vasilakos, Tcp-like molecular communications. IEEE J. Sel. Areas Commun. 32(12), 2354–2367 (2014) 63. I. Isik, How mobility of transmitter and receiver affects the communication quality. AIP Adv. 12(2), 025205 (2022) 64. M. Alam, N. Nurain, S. Tairin, M. Naznin, A.A. Al Islam, Conjugate congestion control based transport layer protocol for molecular communication in body area nanonetworks (BANs), in 2018 5th International Conference on Networking, Systems and Security (NSysS) (IEEE, 2018), pp. 1–6 65. W. Pan, X. Chen, X. Yang, N. Zhao, L. Meng, F.H. Shah, A molecular communication platform based on body area nanonetwork. Nanomaterials 12(4), 722 (2022) 66. M. Alam, N. Nurain, S. Tairin, A.A. Al Islam, Energy-efficient transport layer protocol for hybrid communication in body area nanonetworks, in 2017 IEEE Region 10 Humanitarian Technology Conference (R10-HTC) (IEEE, 2017), pp. 674–677 67. M. Egan, B.C. Akdeniz, B.Q. Tang, Stochastic reaction and diffusion systems in molecular communications: Recent results and open problems. Digit. Signal Process. 124, 103117 (2022) 68. F. Vakilipoor, F. Ratti, H. Awan, M. Magarini, Low complexity receiver design for timevarying Poisson molecular communication channels with memory. Digit. Signal Process. 124, 103187 (2022) 69. L. Chouhan, M.-S. Alouini, Rescaled Brownian motion of molecules and devices in threedimensional multi-user mobile molecular communication systems. IEEE Trans. Wirel. Commun. 21(12), 10472–10488 (2022) 70. S. Ansari, K.A. Alnajjar, Multi-hop genetic-algorithm-optimized routing technique in diffusion-based molecular communication. IEEE Access 11, 22689–22704 (2023) 71. A.-J. Garcia-Sanchez, R. Asorey-Cacheda, J. Garcia-Haro, J.-L. Gomez-Tornero, Dynamic multi-hop routing in terahertz flow-guided nanosensor networks: A reinforcement learning approach. IEEE Sens. J. 23(4), 3408–3422 (2023) 72. T. Furubayashi, T. Nakano, A. Eckford, Y. Okaie, T. Yomo, Packet fragmentation and reassembly in molecular communication. IEEE Trans. Nanobiosci. 15(3), 104 (2016) 73. L. Chouhan, P.K. Upadhyay, P.K. Sharma, A.M. Salhab, On anomalous diffusion of devices in molecular communication network. Preprint (2022). arXiv:2203.14808 74. L.C. Cobo, I.F. Akyildiz, Bacteria-based communication in nanonetworks. Nano Commun. Netw. 1(4), 244–256 (2010) 75. M.J. Moore, T. Nakano, Addressing by beacon distances using molecular communication. Nano Commun. Netw. 2(2-3), 161–173 (2011) 76. L. Hong, W. Chen, F. Liu, Cooperative molecular communication for nanonetwork, in 2014 Sixth International Conference on Ubiquitous and Future Networks (ICUFN) (IEEE, 2014), pp. 369–370 77. M. Kuscu, O.B. Akan, Detection in molecular communications with ligand receptors under molecular interference. Digit. Signal Process. 124, 103186 (2022) 78. Z.P. Li, J. Zhang, T.C. Zhang, Concentration aware routing protocol in molecular communication nanonetworks, in Applied Mechanics and Materials, vol. 556 (Trans Tech Publ., 2014), pp. 5024–5027

60

2 Analysis of Layer’s Tasks in Molecular Communication: Application. . .

79. M.M.U. Rahman, Q.H. Abbasi, N. Chopra, K. Qaraqe, A. Alomainy, Physical layer authentication in nano networks at terahertz frequencies for biomedical applications. IEEE Access 5, 7808–7815 (2017) 80. Y. Li, L. Lin, W. Guo, D. Zhang, K. Yang, Error performance and mutual information for IoNT interface system. IEEE Internet Things J. 9(12), 9831–9842 (2022) 81. A. Rizwan, A. Zoha, R. Zhang, W. Ahmad, K. Arshad, N.A. Ali, A. Alomainy, M.A. Imran, Q.H. Abbasi, A review on the role of nano-communication in future healthcare systems: A big data analytics perspective. IEEE Access 6, 41,903–41,920 (2018) 82. X. Chen, Y. Huang, L.-L. Yang, M. Wen, Generalized molecular-shift keying (GMoSK): Principles and performance analysis. IEEE Trans. Mol. Biol. Multi-Scale Commun. 6(3), 168–183 (2020) 83. S. Senturk, I. Kok, F. Senturk, Internet of nano, bio-nano, biodegradable and ingestible things: A survey. Preprint (2022). arXiv:2202.12409 84. S. Figueiredo, N. Souto, F. Cercas, Low-complexity channel codes for reliable molecular communication via diffusion. Sensors 22(1), 41 (2021) 85. I.F. Akyildiz, M. Ghovanloo, U. Guler, T. Ozkaya-Ahmadov, A.F. Sarioglu, B.D. Unluturk, Panacea: An internet of bio-nanothings application for early detection and mitigation of infectious diseases. IEEE Access (2020) 86. A. Pattanayak, S. Dhal, S.K. Addya, Automatic Privacy-Preserving Contact Tracing of Novel Coronavirus Infection by Cloud-Enabled WBAN Using Blockchain. Cryptology ePrint Archive (2020) 87. M.M. Al-Zubi, A.S. Mohan, P. Plapper, S.H. Ling, Intra-body molecular communication via blood-tissue barrier for internet of bio-nano things. IEEE Internet Things J. 9(21), 21802– 21810 (2022) 88. L. Kong, L. Huang, L. Lin, Z. Zheng, Y. Li, Q. Wang, G. Liu, A survey for possible technologies of micro/nanomachines used for molecular communication within 6G application scenarios. IEEE Internet Things J. (2023) 89. S. Ghavami, R. Adve, F. Lahouti, Information rate analysis of ask-based molecular communication systems with feedback. Nano Commun. Netw. 28, 100339 (2021) 90. J. Wang, M. Peng, Y. Liu, X. Liu, M. Daneshmand, Performance analysis of signal detection for amplify-and-forward relay in diffusion-based molecular communication systems. IEEE Internet Things J. 7(2), 1401–1412 (2019) 91. M. Amjad, M. Lestas, H.K. Qureshi, T. Saeed, A. Pitsillides, Concentration gradient based routing for molecular nanonetworks, in 2020 International Conference on Information Networking (ICOIN) (IEEE, 2020), pp. 336–341 92. F. Walsh, S. Balasubramaniam, Reliability and delay analysis of multihop virus-based nanonetworks. IEEE Trans. Nanotechnol. 12(5), 674–684 (2013) 93. S. Balasubramaniam et al., Opportunistic routing through conjugation in bacteria communication nanonetwork. Nano Commun. Netw. 3(1), 36–45 (2012) 94. S. Balasubramaniam et al., Multi-hop conjugation based bacteria nanonetworks. IEEE Trans. Nanobiosci. 12(1), 47–59 (2013) 95. A. Aijaz, Opportunistic routing in diffusion-based molecular nanonetworks. IEEE Wirel. Commun. Lett. 4(3), 321–324 (2015) 96. D.C. Ferreira, L.P. Reis, N.V. Lopes, A nanocommunication system for endocrine diseases. Clust. Comput. 20(1), 689–706 (2017) 97. H. Byun, Feedback-controlled adaptive signal detection scheme for diffusion-based molecular communication systems. Appl. Sci. 13(4), 2171 (2023) 98. M. Stelzner, F. Dressler, S. Fischer, Function centric nano-networking: Addressing nano machines in a medical application scenario. Nano Commun. Netw. 14, 29–39 (2017) 99. A. Einolghozati, M. Sardari, A. Beirami, F. Fekri, Consensus problem under diffusion-based molecular communication, in 2011 45th Annual Conference on Information Sciences and Systems (IEEE, 2011), pp. 1–6

References

61

100. G. Yue, Q. Liu, Cooperative target tracking algorithm based on massive Beacon coordinates system in directional molecular communication. IEEE Trans. NanoBiosci. 21(3), 405–415 (2022) 101. T. Kotsuka, Y. Hori, Frequency response of diffusion-based molecular communication channels in bounded environment. Preprint (2022). arXiv:2203.13532 102. R. Paridar, N. Mokari, E. Jorswieck, M.R. Javan, On the reception process of molecular communication-based drug delivery. Preprint (2021). arXiv:2112.12485 103. Z. Cheng, J. Sun, J. Yan, Y. Tu, Optimizations for mobile mimo relay molecular communication via diffusion with network coding. KSII Trans. Internet Inf. Syst. (TIIS) 16(4), 1373–1391 (2022) 104. A. El-atty, M. Saied, R. Bidar, E.-S.M. El-Rabaie, Embedded nano relay for intra-body network-based molecular communications. Wirel. Pers. Commun., 1–18 (2022) 105. X. Liao, Z. Jia, Z. Wu, B. Liu, X. Wang, X. Jiang, Modulation and signal detection for diffusive-drift molecular communication with a mobile receiver. Mobile Inf. Syst. 2021, 1–17 (2021) 106. I. Iqbal, M. Nazir, A. Sabah, Design of energy-efficient protocol stack for nanocommunication using Greedy Algorithms. J. Comput. Netw. Commun. 2022, (2022) 107. I.F. Akyildiz, J.M. Jornet, M. Pierobon, Nanonetworks: A new frontier in communications. Commun. ACM 54(11), 84–89 (2011) 108. L. Donvito, L. Galluccio, A. Lombardo, G. Morabito, mu-net: A network for molecular biology applications in microfluidic chips. IEEE/ACM Trans. Netw. 24(4), 2525–2538 (2015) 109. M.C. Gursoy, A.E. Pusane, T. Tugcu, Molecule-as-a-frame: A frame based communication approach for nanonetworks. Nano Commun. Netw. 16, 45–59 (2018) 110. X. Chen, M. Wen, F. Ji, Y. Huang, Y. Tang, A.W. Eckford, Detection interval for diffusion molecular communication: How long is enough? Preprint (2022). arXiv:2204.08636

Chapter 3

Analysis of the Molecular Physical Layer’s Tasks

In this chapter, the molecular physical layer will be approached in a very detailed way. First, we will analyze the physical fundamentals to produce the communication of molecules in diverse scenarios, and subsequently, the task of this layer will be discussed. The reason to investigate the physical characteristic of a molecular communications (MCs) system deeply is related to the importance that MCs have in science and, thus, in many technical research paradigms. Since health is one of the most essential topics, academics from many disciplines are working hard to make the current healthcare system better. Information and communication technologies (ICT) have lately been employed in the field of healthcare to create intelligent systems that speed up the process of abnormality detection in human bodies and provide an accurate and efficient treatment. The Internet of Medical Things (IoMT) has automated healthcare system processes ranging from data gathering to diagnosis [1]. The Internet of Things (IoT) devices used in medical healthcare systems are expressed by the acronym IoMT. Consequently, throughout this investigation, we will utilize both terminologies equally. The old healthcare systems have been transformed by IoMT into smart healthcare systems [2, 3]. As demonstrated in this book, MCs are essential for future medical applications, such as the identification, evaluation, and management of infectious disease epidemics. The COVID-19 pandemic is endangering the health and financial security of the entire world’s population. A significant difficulty for medical care is how to provide quick disease identification combined with “on-site” results because of its rapid dissemination. In this regard, it was suggested to implement remote health diagnosis and monitoring during COVID-19 using IoMT, an application and a specialization of IoT in the medical sector. The IoMT network typically consists of body sensors, local devices, and edge servers, with the sensor acting as the network’s foundation because it allows for the collection of various health signals for the sake of a subsequent health evaluation [4].

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 Y. Cevallos et al., Molecular Communications, https://doi.org/10.1007/978-3-031-36882-0_3

63

64

3 Analysis of the Molecular Physical Layer’s Tasks

IoMT sensors collect patient data such as blood pressure, oxygen levels, glucose levels, and pulse rate and transmit it to doctors, patients’ family members, or the closest caretakers in these smart systems. The time that would have been spent manually recording a patient’s body parameters has been greatly reduced because of automation in the healthcare industry. A potent technology during the COVID-19 pandemic was IoMT. It made it possible to diagnose, monitor, and treat patients remotely, which helped stop the virus’s spread. To quickly gather more information and statistics about the infection, data from various patients is automatically gathered, processed, and stored in the cloud. Also, it made healthcare accessible at any time and in any place. It made it possible for nurses to attend to multiple patients at once, saving countless lives. The traditional healthcare systems, which require a caretaker for every patient and are unable to handle the increasing number of patients, will be much improved by the IoMT-enabled smart healthcare system [2]. Nanobiosensors are thought to be a viable technology for building it but are limited by the size and biocompatibility of sensors. As part of the Internet of Bio-Nano Things, researchers also suggested employing nanobiosensors to make it easier to identify and treat a variety of diseases (IoBNT). At this stage, the IoBNT is a unique realization of the IoMT in every person [4]. The connection of sensors positioned in various tissues is a critical problem for IoBNT or IoMT to be resolved in order to collect all monitored medical data from patients. A number of communication technologies, including molecular communication (MC), electromagnetic communication, and acoustic communication, have been proposed in the literature for use with biological nanosensors. MC is a bio-inspired communication paradigm that differs from traditional communication channels and uses chemical signals as its information carrier [5]. Accordingly, MC has the following properties: small size, energy efficiency, and excellent biocompatibility [6]. Considering the aforementioned benefits, MC may be able to link larger groups of nanobiosensors in IoBNT [4]. On the other hand, MC can assist the biomedical area in applications like targeted drug delivery [7], disease diagnostics, and health monitoring as an interdisciplinary communication method. Also, MC may be used as a powerful tool to predict the transmission of infections and diseases via aerosols during the COVID-19 outbreak or later pandemics. In light of this, MC can open the door for the IoMT to link sensors everywhere over the body [4]. Another medical condition that can be examined from MC’s point of view is thrombosis, which kills one in four people worldwide and is one of the primary causes of death. A blood clot develops in a vein during thrombosis; if it does, it can prevent or delay blood flow or be detected in organs. A new IoBNT-based model that predicts and analyzes blood vessel coagulation has been developed with a bio-cyber communication interface. This paradigm allows for the collection of information from the blood vessel, which is then converted into electrical equivalent using the bio-cyber interface. Moreover, the release of particular nanocarrier molecules like liposomes, nanodevices that may travel through the bloodstream and foretell clots, has been stimulated by the optical or thermal response [8].

3 Analysis of the Molecular Physical Layer’s Tasks

65

IoBNT can potentially be utilized to detect and treat infectious infections early. Examples include the development of a network for cystic fibrosis; the genetic illness of insulin and glucose concentration that manifests in organs, such as the pancreas, intestines, lungs, and sweat glands; and the direct transmission of values to medical professionals. In this way, pump life might emerge in waves and result in patient death in addition to monitoring health status. The idea of an extension by ensuring that the insulin cartridge lasts longer is mentioned by authors in [8]. It cites an application with an IoBNT network named PANACEA as well as the protection the body receives from such intelligent systems, which offers an end-to-end remedy to infectious disorders, and from the negative effects of too much insulin. On the other hand, in this network, submillimeter implantable bioelectronic preventive health services can be offered by means of devices like those that sense communication within body cells and are used as pre-illness therapy with a holistic approach to physical and, to assess the level of infection, psychological diseases. In this situation, it is expected that making changes will enable the creation of healthy living conditions. An application for people’s lifestyles with IoBNT centered on modeling IoBNT applications that will improve the diagnosis, management, and treatment practices of the insulin-glucose system for this condition is then described in [8] as another prevalent disease nowadays. The invention of several materials, including medical implants, films, environmental sensors, non-injectable sensors, edible and biodegradable sensors, films, and disposable devices, has also been made possible by recent developments in synthetic chemistry and engineering. Electronics that are biodegradable and editable function for a while before being destroyed by hydrolysis or biological processes [9, 10]. These devices, in their intelligent forms, can be employed in a variety of industries, including the production of medical equipment, drug delivery, tissue engineering, and food packaging. It is now possible to use the various types of electrical and biomaterials outlined above for the benefit of the environment and people because of advancements in biology, materials science, and engineering. Now, two distinct subtypes of IoBNT, the Internet of Biodegradable Things (IoBDT) and the Internet of Ingestible Things (IoIT), are emerging. These networks can be created using IoNT- and IoBNT-specific components and architectural frameworks [11]. The edge devices or sensors used in these networks are different, though, and they are comprised of edible, digestible, and biodegradable sensors or devices that are application-specific [8]. However, the developed electronic sensors must be capable of quick responses, and after usage, biodegradation and waste conditions must be managed. In order to find stomach tumors and ulcers, for instance, ingestible chemical sensors record changes and pH levels in the gastrointestinal tract. Moreover, a wide pH range is presented to oral drugs. As a result, drug components may decay at the wrong periods and lose some of their effectiveness. But smart pills also have medication release processes, feedback algorithms, and environmental sensors. The goal of some researchers is to create a temperature sensor with an ultrathin format, a quick response time (10 ms), and a stable operation with minimal resistance change

66

3 Analysis of the Molecular Physical Layer’s Tasks

when the device is folded. As a result of these research efforts, a sensor with high mechanical stability and biodegradability has been created. An ingestible hydrogel device with a quick transition feature that may be ingested as a pill and take the shape of a soft spherical was introduced in another significant study [8]. The previous scenery alludes to communication inside the human body, i.e., in the field of action of molecular communications. Thus, it is crucial to understand how this communication is produced [12]. Hence, in this chapter, we focus on the physical layer (i.e., fundamental definitions, propagation, and processing) of information at the molecular level. Subsequently, the network’s functionality in this layer will be analyzed.

3.1 Definition of Nanomachine and Bio-nanomachine A nanomachine is a device with all of its dimensions being at the nanoscale or microscale, and it can carry out operations including computation, data storage, communication, sensing, and actuation [13]. The range of applications in complexity and operations is increased by effective cooperation and coordination among nanomachines. Chemical sensors, nanovalves, nanoswitches, or molecular elevators are examples of nanomachines that are unable to do complex tasks on their own. Networked nanomachines will be able to cooperate and operate in synchrony to carry out more difficult tasks thanks to the exchange of data and commands [14– 25]. A nanomachine may have one or more parts, leading to various degrees of complexity, ranging from straightforward molecular switches to nanorobots. The following architectural elements will be present in the most comprehensive nanomachines [14–24]: 1. Control unit. It aims to carry out the commands to carry out the planned tasks. It has the ability to command every other part of the nanomachine in order to do this. The control unit could come with a storage device where the nanomachine’s data is kept [14–23]. 2. Communication unit. It includes a transceiver that can send and receive messages at the nanoscale, like molecules [14–23]. 3. Reproduction unit. This unit’s job is to reproduce the nanomachine by first fabricating each component from scratch using materials from outside the system. All the instructions required to complete this activity are included in this unit [14–23]. 4. Power unit. The entire nanomachine is intended to be powered by this unit. The device will be able to capture energy from ambient factors like light and temperature and store it for later distribution and use [14–23]. 5. Sensor and actuators. These parts function as an interface between the environment and the nanomachine, just like the communication unit does. A nanoma-

3.1 Definition of Nanomachine and Bio-nanomachine

67

chine may have a number of sensors and/or actuators, such as temperature sensors, chemical sensors, clamps, pumps, motors, or locomotion devices [14– 23]. A biological nanomachine in MC is a nanomachine with biological components (bio-nanomachine1 ). In an aquatic environment, bio-nanomachines exchange molecules to communicate and work together to complete task-oriented tasks. Examples of bio-nanomachines include protein motors that have been rebuilt to move molecules and nanoscale molecular complexes like DNA (deoxyribonucleic acid) molecules that are engineered to carry out logical processes. Microscale, genetically modified cells with environmental sensing capabilities are another example of bio-nanomachines. A collection of bio-nanomachines can communicate and carry out tasks that may not be possible for individual bio-nanomachines to carry out thanks to molecular communication. In a range of application sectors, including nanomedical applications, molecular communication is anticipated to play a significant role [6, 14–22]. Regarding bio-nanomachines, there are three crucial factors to consider. Initially, biological components (such as proteins, nucleic acids, lipids, and living cells) may or may not be present in a bio-nanomachine (e.g., magnetic particles). Second, a bio-nanomachine can be as small as a macromolecule or as large as a biological cell. Under this definition, biological cells are included, which are often far larger than what is commonly meant by the term “nano” (i.e., dimensions of 1–100 nm). Finally, a bio-nanomachine implements a set of basic features, such as basic energy acquisition and expenditure, basic protein filament motion, and basic molecule processing (e.g., capturing, storing, and releasing molecules, detecting molecules, and changing molecules). Biological devices include, for instance [6, 14–22, 26]: • A DNA sequence that can use enzymes to cut and release a segment of a complementary or partially complementary DNA sequence it finds in the environment [6] • A protein motor with the ability to attach to a certain kind of molecule, move along protein filaments transporting the molecule, and unbind the molecule [6] • A liposome that can take in and release specific kinds of molecules [6] • A genetically modified cell that can determine whether the concentration of a particular kind of chemical in its environment falls within a defined range [6] • A biological cell that has been functionalized with nonbiological components like magnetic particles and photosensitive polymers [6]

1 In

Sect. 3.3.6.1.1, this topic will be analyzed from the perspective of tasks in bio-nanomachine sublayer.

68

3 Analysis of the Molecular Physical Layer’s Tasks

3.2 Definition of Nanonetwork (Bio-nanonetwork) A brand-new distributed computer network that works at the nanoscale is called a nanonetwork. The development of communication protocols in nanonetworks is hampered by the simplicity of a single nanomachine and the new communication paradigm. The academic community has access to a wide range of new research areas because of the rapid growth of nanotechnology. For instance, nanonetworks are bringing human intelligence to the microscopic realm. A distributed nanonetwork, made up of multiple nanomachines, should be able to conduct more complicated activities across a greater area because a single nanomachine can only carry out extremely simple tasks. As a result, data interchange between the nanomachines is necessary. Current works on nanonetworks, which are still in the early stages of study, primarily concentrate on the elements and supporting technologies of network architecture. High-performance communication protocols are required for the development of a trustworthy and useful nanonetwork [13, 26]. Single cells can detect chemical concentrations with remarkable accuracy. Its accuracy sometimes comes close to the physical bounds established by molecular diffusion. Yet, no one cell carries out this sensory function alone. Communities of cells exist in tissues, biofilms, and colonies, i.e., they function as a network [27, 28]. Cells in these communities communicate in a variety of ways. Juxtacrine signaling, which is the interchange of molecules between cells in close proximity, and the release and detection of diffusible molecules at distances at least as great as those of a cell serve as communication mechanisms (autocrine signaling). This begs the question of whether cell-cell communication enhances sensory accuracy above and beyond what a cell can accomplish on its own. Cells are more responsive in groups than they are alone, according to experiments. Biased morphological or motile responses are seen in groups of neurons, lymphocytes, and epithelial cells in response to chemical gradients that are too shallow for individual cells to detect. In fruit fly embryos, groups of cell nuclei detect morphogen concentrations with more accuracy than would be predicted for a single nucleus. Cell-cell communication has been demonstrated in some circumstances to be directly responsible for increased sensitivity, such as with epithelial cells. Yet, from a theoretical standpoint, concentration sensing and gradient sensing have usually been constrained to single receptors or single cells. Only for specific geometries have analogous limits for networks of communicating cells been determined, and they are otherwise poorly understood. It is particularly unclear how the restrictions scale with collective characteristics like communication intensity and population size as well as whether they depend on the communication method (juxtacrine versus autocrine) [27, 29]. The referred concepts are essential in MC research, industry, and academia. They match with these defined by the IEEE (Institute of Electrical and Electronic Engineers) through two standards in nanocommunication (for wireless and molecular level), i.e., IEEE 1906.1 [30] and IEEE 1906.1.1 [31]. These standards define functions of “molecular layers” in a molecular network stack (analyzed in Chap. 4) in which it is established that communication is the act

3.3 Molecular Communication Systems

69

of conveying a message from a transmitting party to a receiving party. This includes the components of the message, transmitter, receiver, medium, and message carriers. In nanoscale communication at one to a few nanometers (nm), communication at the nanolevel includes systems with many transmitters and receivers, such as broadcast, multicast, and network communication systems. The definition of a message comprises signals transmitted for control purposes. Nanoscale communication networks shall describe their physical layer by denoting the transmitter, receiver, message, medium, and components that have a dimension from 1 to 100 nm, the communication physics [30, 31].

3.3 Molecular Communication Systems Building the theoretical underpinnings of molecular communication requires the use of biological molecules in communication technology, namely, molecular communication. Since it has made a substantial contribution to the development of the present telecommunications technology, Shannon’s information theory may be relevant. Developing molecular codes, measuring and assessing channel capacity, and coming up with effective design strategies for molecular communication in the presence of thermal noise are some of the general research issues. In the following paragraphs, these topics will be discussed from the ICT (Information and Communication Technology) perspective, and then MCs will be decomposed into the basic components of conventional communication systems [25, 32]: 1. Communication Components. Whereas telecommunication uses silicon-based electric devices for communication, molecular communication enables the networking of nanomachines. In molecular communication, nanomachines are nanoscale to microscale devices found in biological systems or produced synthetically from biological components [25, 32]. 2. Signal Types. Contrary to communication technology, which employs electrical or optical impulses, molecular communication uses chemical signals to transmit information. The use of signal molecules as information carriers creates new opportunities for ICT [33–39]. Consider the physical characteristics that signal molecules possess, which are highly information dense. Moreover, signal molecules might be functional. For instance, a nanomachine that receives a DNA sequence that codes for some biological processes may develop new functionality (such as a functioning protein) as a result of gene expression. 3. Communication Media. Whereas electrical impulses in telecommunication (or electromagnetic waves) go through a metallic wire, chemical signals in molecular communication travel through an aquatic environment (or in the air). Thermal noise in a communication medium during molecular communication has a substantial impact on how signal molecules spread throughout the communication medium. Moreover, a communication medium contains other molecules that could interact with signal molecules and impair communication [32].

70

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.1 Communication channels

A communication pathway gives signal molecules a way to go between cells. The passive and active modes of propagation are depicted in Fig. 3.1. Signal molecules diffuse at random in the passive mode (Fig. 3.1a) due to Brownian motion [40, 41]. Although no external mechanism is required, a large number of molecules or a long enough time must be present for signal molecules to travel a great distance. Furthermore, in high-viscosity environments, big signal molecules propagate very slowly. On the other hand, signal molecules are directly propagated in the energy-consuming active mode (Fig. 3.1b) (e.g., via ATP hydrolysis). Active propagation may shorten the amount of time signal molecules must travel to the receiver while also increasing the likelihood that they will do so. Yet, these procedures frequently need chemical energy and guiding and transport molecules for communication [32]. Signal molecule propagation has been developed and designed using a variety of ways. For instance, a line of biological cells with their interiors directly linked by cell-to-cell communication pathways has been used to create a passive propagation mechanism. Signal molecules travel along the cell wire, which is a line of cells. The technique can effectively spread signal molecules between cells since there is no diffusion to the surrounding environment and signal molecule diffusion is limited to the interiors of connected cells. The addition of an amplification process during propagation has also improved passive propagation mechanisms. In this

3.3 Molecular Communication Systems

71

instance, signal molecules move in a way like waves through repeated diffusion and amplification processes [3, 32, 42–44]. Moreover, active propagation techniques have been developed. One illustration of such a process makes use of molecular motors, which serve as transport molecules and actively move signal molecules along the guide molecules, as well as molecular rails, which serve as guide molecules and self-organize into a network. The active transport mechanisms in biological cells and this mechanism are comparable. Molecular motors can be selectively propagated along a defined topology to specific sites on a network by directing self-organization processes to construct a topology of molecular railways. In another artificial active propagation mechanism, the molecular motors that serve as the guide molecules are coated on a glass surface, inverting the position of the microtubules and the motors. The molecular rails that travel along the surface are pushed along by the motors. In this configuration, signal molecules are loaded onto one nanomachine by the filaments and discharged onto a different one. As a result, the filaments serve as transport molecules. It is possible to design several molecular motor designs that draw transport molecules to a particular nanomachine [32]. These types of transport will be discussed in Sect. 3.3.2.

3.3.1 Molecular Communication System Components As it has been analyzed in the previous paragraphs, due to their biological nature, MC systems consist of the three fundamental components same as in conventional communication systems, i.e., transmitter, communication channel, and receiver. A fundamental method of molecular communication is shown in Fig. 3.2. According

Fig. 3.2 Molecular communication components

72

3 Analysis of the Molecular Physical Layer’s Tasks

to this paradigm, signal molecules are transmitted between sender and receiver nanomachines across a communication channel. In n-to-n communication, the transmitter and receiver frequently represent a collection of molecular systems. A sender (or senders) transmits the signal molecules, which then travel along the communication channel and are received by the recipient(s) of the message. While the communication channel frequently contains noise sources like thermal noise and other molecules that prevent signal molecules from propagating, it also provides a way for signal molecules to do so [32, 42, 43]. Transmitters and receivers deal with the emission and reception of molecular information, respectively, and are different from typical information signals; in MCs, there is no existing information carrier to facilitate its propagation, and then the molecular information is physically transported by itself through physical and chemical reactions [32, 42, 43]. When communicating, the transmitter modifies its surroundings, and the receiver needs to be able to measure those changes. A wireless transmitter causes a changing EM (electromagnetic) field along an antenna, which may be detected in an antenna at the receiver. Once more, this is true of any communication system. In contrast, the change in molecular communication must be molecular: the transmitter releases molecules into a shared channel, which travel to (and are detected by) the receiver [43, 45–48]. Regarding signal molecules, some examples that take part in cell communication include endocrine hormones, local mediators (e.g., cytokines), neurotransmitters (e.g., dopamine, histamine), and intracellular messengers (e.g., .Ca 2+ and cyclic AMP). DNA and RNA (ribonucleic acid) molecules are also signal molecules that store and transfer genetic information in the cell [32, 42, 43]. There are some methods and possibilities for facilitating nanomachine communication (Fig. 3.3). The first choice is to designate the recipient of a signal molecule using an addressing mechanism. DNA sequences can be used to construct a generic addressing method. A signal molecule is joined to a single-stranded DNA molecule that contains the address of a receiver [32, 42, 43].

Fig. 3.3 Signal molecules

3.3 Molecular Communication Systems

73

The receiver bio-nanomachines chemically react with the propagating molecules during molecular communication, which involves sending information molecules from a group of sender bio-nanomachines to a group of receiver bio-nanomachines. The physical or chemical characteristics of molecules, such as their type, their three-dimensional structure (e.g., protein structure), sequence structure (e.g., DNA sequence), or concentration (e.g., calcium concentration) that the recipient bionanomachines can react to, are used to represent information in molecular communication. In a molecular structure, a great density of information may be encoded. For instance, if a DNA base is one of the four DNA bases (A, T, G, or C) and measures about 0.34 nm in length, a DNA sequence may be able to store two bits of information within that length. Functional information can also be encoded. For instance, a bio-nanomachine that receives a DNA sequence encoding a particular protein may also acquire new capabilities (such as tolerance to hazardous chemicals) as a result of gene expression [43, 45, 49]. When the signal molecule reaches the designated receiver, the receiver forms double-stranded DNA because it possesses a DNA sequence that is complementary to the single-stranded DNA on the signal molecule. Given that a DNA sequence of 10.000 base pairs can be created using current DNA technology, the DNAbased addressing system has enough of address space. The second choice is to enclose a signal molecule in an interface molecule. Regardless of the sorts of signal molecules inside, an interface molecule offers a general abstraction for communication, enabling nanomachines to interchange a range of molecules utilizing the same communication method. A vesicle or spherical lipid bilayer resembling the membrane covering a real cell can be used to implement such an interface molecule. Moreover, an interface molecule can stop signal molecules from propagating and having unwanted interactions with other molecules in the environment [43, 45, 49]. Additionally, for end-to-end communication between nanomachines at the destination and a signal from outside the environment, these nanomachines must receive (detect), process, and send signal molecules. Figure 3.4 depicts a sample nanomachine design along with several methods and alternatives for taking in, analyzing, and transmitting signal molecules. A nanomachine collects signal molecules from the environment during receiving (sensing). As no specific mechanism is required for their uptake if the signal molecules are membrane-permeable, they can act directly on the interior of the nanomachine. Other solutions are available in various circumstances. The use of surface receptor molecules (also known as membrane receptors), which bind the signal molecules and cause signal transduction in the nanomachine, is one such possibility. Another option is to let signal molecules flow into the nanomachine by opening surface channels (such as membrane-bound channels). Moreover, the outer membrane of a nanomachine may fuse with an interface molecule (such as a vesicle) in order to convey signal molecules contained inside into the interior of the nanomachine [43, 45, 49]. In processing, signal molecules are biochemically reacted with by a nanomachine, altering its chemical state. Certain molecules need to be identified, created, stored, or destroyed by a nanomachine as part of the processing mechanism. A

74

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.4 Nanomachine architecture on the cell

nanomachine’s processing of molecules may result in the release of molecules as well as mechanical activity that causes division, migration, or aggregation [43, 45, 49]. A nanomachine sends by releasing signal molecules into the communication path (or the environment). Similar to transmitting, membrane-permeable signal molecules can be sent without the aid of a mechanism since they can diffuse across the outer membrane [43, 45, 49]. In other situations, a variety of methods are available for transferring signal molecules from a nanomachine’s inner to the outside world. For instance, a nanomachine might release any signal molecules bound to the surface receptor molecule and release them into the surrounding space. Moreover, a nanomachine may release signal molecules by widening surface channels (or by utilizing pumps), allowing the molecules to disperse. Furthermore, by budding an interface molecule (such as a vesicle) containing the signal molecules, a nanomachine may release signal molecules to the environment outside [43, 45, 49]. The fact that biological or MC communication is regarded as stochastic is a distinguishing trait in these types of communication. Shannon’s communication

3.3 Molecular Communication Systems

75

Fig. 3.5 Stochastic model of molecular communication

model can be used to define the fundamental model of molecular communication (Fig. 3.5). It is made up of elements that act as information molecules that stand in for the information (or messages) that need to be transmitted, biological senders that send the information molecules, biological receivers that receive the information molecules, and the environment in which the information molecules travel from the biological sender to the biological receiver. Other types of specialized biological components may also be present, such as addressing molecules (not shown), which are attached to information molecules or interface molecules to specify the biological receiver; guide molecules, which direct the movement of transport molecules; and interface molecules, which enable a transport molecule to transport information molecules selectively [40, 50]. Figure 3.5 shows a generic molecular communication architecture. It consists of components functioning as information molecules that represent the information to be transmitted (e.g., endocrine hormones, local mediators such as cytokines, neurotransmitters such as dopamine and histamine, intracellular messengers such as .Ca 2+ and cyclic AMP, DNA/RNA molecules, and synthetic molecules such as nanoparticles), sender nanomachines that release the information molecules (biological cells, genetically engineered cells, and artificial cells), receiver nanomachines that detect information molecules, and the environment in which the information molecules propagate from the sender nanomachine to the receiver nanomachine. It may also include transport molecules (e.g., molecular motors) that transport information molecules, guide molecules (e.g., microtubule or actin filaments) that direct the movement of transport molecules, and interface molecules (e.g., vesicles)

76

3 Analysis of the Molecular Physical Layer’s Tasks

that allow any information molecule to be transported by a transport molecule [43, 45, 49]. The molecular communication components carry out the following five phases of communication: encoding information into an information molecule by the sender nanomachine, sending the information molecule into the environment, propagating the information molecule through the environment, receiving the information molecule by the receiver nanomachine, and decoding the information molecule into a chemical reaction at the receiver nanomachine [43, 45, 49]. A sender nanomachine converts information into molecules that the receiver nanomachine can detect during the encoding phase. Information can be stored in information molecules by encoding it in a variety of ways, such as the information molecule’s three-dimensional shape, the specific molecules that make up the information molecule, or the concentration of information molecules (the number of information molecules per unit volume of solvent molecules) [43, 45, 49]. A sender nanomachine distributes information molecules into the environment during the sending phase. If a sender nanomachine is a biological cell with endoplasmic reticulum, for example, the information molecules can be unbound from the sender nanomachine by budding vesicles from the biological cell, or a gate can be opened to let the information molecule diffuse away (e.g., by opening a gap junction channel in the cell membrane of a sender nanomachine). Another possibility for a sender nanomachine is to catalyze a chemical reaction that results in transport molecules [43, 45, 49]. The phase of propagation is when the information molecule travels from the sender nanomachine to the receiver nanomachine through the surrounding environment. Either the information molecule may bind to a transport molecule (such as a molecular motor that creates motion) and actively propagate across the environment (lower panel in Fig. 3.5) using chemical energy or it may diffuse passively (upper panel in Fig. 3.5) without using chemical energy. During propagation, an interface molecule can also be required to shield information molecules from outside noise. A vesicle-based interface molecule, for example, may include information molecules that spread throughout the surroundings. The vesicle stops the information molecules from interacting chemically with molecules outside of it [43, 45, 49]. During the receiving phase, the receiver nanomachine absorbs information molecules that are spreading throughout the surroundings [43, 45, 49]. Following the capture of information molecules by the receiver nanomachine, the decoding phase is when the molecules are decoded into a chemical reaction. Using receptors that can bind to a particular kind of information molecule is one way to capture information molecules. Using channels (such as gap junction channels) that enable them to flow into a receiver nanomachine without the need for receptors is an additional method of catching them. At the receiver nanomachine, chemical processes for decoding can involve creating new molecules, carrying out a straightforward operation, or creating another signal (e.g., sending other information molecules) [43, 45, 49].

3.3 Molecular Communication Systems

77

Fig. 3.6 Molecular communication system model

Shannon’s paradigm applied to MC systems establishes more realistic components having in mind the stochastic behavior (particularly in molecular communication channels); this scenery is exhibited in Fig. 3.6 in which the components of MC systems are described as: 1. Bio-nanomachines. To create molecular communication systems, these sensors and actuators are used. Material, size, and functionality make up the three characteristics of a bio-nanomachine. A bio-nanomachine is made out of biomaterials, either alone or in combination with nonbiological components. A bio-nanomachine can be as small as a few tens of m, which is the size of a biological cell or a macromolecule. A bio-nanomachine uses a variety of biological functions, such as sensing a certain kind of molecule, to accomplish a variety of application-specific objectives (e.g., biomolecular sensing). Examples of bio-nanomachines are genetically engineered biological cells [49, 51]. In a molecular communication environment, where molecular communication occurs, reside bio-nanomachines. Bio-nanomachines can live in, communicate with, and make noise in the molecular communication environment. Signal propagation in varied media and situations is one of the most crucial problems in physical communication in MCs [6]. 2. The Environment (Communication Channel). The setting for the deployment of bio-nanomachines is a three-dimensional one. For bio-nanomachines to function, an aqueous environment with molecules and energy sources is often required (such as the interior of the human body). The environment might potentially

78

3 Analysis of the Molecular Physical Layer’s Tasks

include flow that could impede or facilitate the operation of biological machinery. The environment may produce interesting occurrences or information, or it may contain objects that produce it. For instance, a transition in the human body into an aberrant condition reflects such an event in biomolecular sensing applications. Cancer cells are a target in applications for targeted medication delivery [49, 51]. The molecular communication environment typically operates at the nanoscale to microscale, providing energy sources for bio-nanomachines to use (such as adenosine triphosphate (ATP) molecules) and an aqueous environment with suitable biochemical conditions for their survival and operation (such as a suitable temperature and pH level) [6]. The molecular communication environment offers space or fluid as the communication medium, and molecules may passively move through the medium through stochastic diffusion in the environment or by being carried by the fluid in the environment (e.g., a blood artery). Passive transport is the name given to this method of molecule propagation (Fig. 3.7a). This form of environment is usual for molecular communication, and it takes little work to set up this kind of medium in the environment [6]. Moreover, a network of protein filaments or concentration gradients of particular molecules in the environment could serve as the communication medium that the molecular communication environment offers, i.e., using mobile bio-nanomachines as a means of active molecule propagation in a particular direction (e.g., by using protein motors carrying

Fig. 3.7 (a) Passive transport and (b) active transport of molecules in the molecular communication environment

3.3 Molecular Communication Systems

79

molecules over a network of protein filaments or by using self-propelling organisms carrying molecules through the environment). Active transport is the name given to this method of molecule transmission (Fig. 3.7b). Connectivity between bio-nanomachines (such as a network of protein filaments and concentration gradients of molecules that self-propelling organisms follow) must be established in the environment beforehand either artificially or through the use of autonomous behaviors in order to use this type of communication medium [6, 52]. The development of the species over time has been made possible via molecular communication; however, as a communication system, they also come with drawbacks like noise [6]. 3. Sources. These are environmental elements that offer helpful information. A bio-nanomachine or a set of bio-nanomachines that detect an event or a target in the environment might be considered a source. When acting as a source, a bio-nanomachine can detect an event or target biochemically and transfer the information by propagating information molecules to sinks or other sources [49, 51]. 4. Sinks. These are organizations that receive and gather data from biological machines or other sources. A sink could be a bio-nanomachine or a collection of bio-nanomachines that analyze data and carry out activities that are specific to an application, like releasing medication molecules. A sink could also be a typical device with traditional communication capabilities (e.g., wireless communication). A sink might be several orders of magnitude bigger than bio-nanomachines and built of environmentally unfriendly materials. When connecting molecular communication systems with external systems, a sink may act as a gateway. Implantable medical devices are an example of a sink device [6, 49, 51]. 5. Noise. Several kinds of this contamination can be found in molecular communication networks. Thermal noise is the first category of noise. Thermal noise causes molecular communication between bio-nanomachines and their operation to be stochastic. Physical noise is the second category of noise. It is challenging for molecules to spread and for bio-nanomachines to move due to the high viscosity of the environment and fluid in the environment. The third type of noise is brought on by either noise molecules or molecules found in the environment. Noise molecules can disrupt the chemical communication between bio-nanomachines and their functionality [49, 51]. A molecular communication environment contains molecules that may naturally occur in the environment because it is an aqueous environment with specific biochemical conditions (at a particular temperature and pH level) (e.g., .Ca 2+ in a biological cell), may be artificially introduced into the environment (e.g., drug molecules in a biological cell in a target drug delivery application [53–58], may be created through bio-nanomachines and energy sources decomposing in the environment, or may be created as a result of biochemical reactions among molecules in the environment. As a result, the molecular communication environment creates disturbances (i.e., noise) to bio-nanomachines and molecules that bio-nanomachines use to communicate [6]. Thus, some types of noise found in the molecular communication environment include:

80

3 Analysis of the Molecular Physical Layer’s Tasks

• Biochemical noise. Biochemical environmental circumstances and environmental chemicals may interact biochemically with biological machineries and the molecules they utilize to communicate. • Thermal noise. Varying environmental temperatures result in varying levels of thermally activated process activity and introduce stochastic thermal motion of molecules and bio-nanomachines that are used for communication [6]. • Physical noise. Bio-nanomachines and the molecules they employ for communication find it challenging to move in a particular direction due to the environment’s physical force-creating fluid and viscosity [6]. Moreover, not all particles that are discharged might reach the receiver. This might be as a result of the particle degradation or chemical interactions with other particles, as well as the transmission scheme. The process of reception and detection may also be noisy. In the biological context, a typical receptor structure is stereochemically matched to a specific signaling particle or molecule, and both the quantity and density of receptors and the kinetics of the binding and unbinding process must be taken into account [59]. The poor speed of MCs results in additional negative impacts in addition to noise in an MC channel. This kind of “cross-talk” or blatant interference must be taken into account when thinking about networks of molecular transceivers. The intersymbol interference (ISI) issue is a significant barrier in MC systems. During the course of each symbol period, the transmitter emits particles, some of which may linger in the atmosphere and disrupt upcoming broadcasts. As more particles from earlier transmissions accumulate in the channel, this issue gets worse with time. ISI mitigation approaches are based on signal processing, using various types of particles and switching between them during each channel use and inserting enzymes in the propagation environment to degrade the information particles in order to address this issue [49, 59, 60]. These techniques will be investigated in Sect. 3.3.5 corresponding to molecular modulation, and error control for noise and ISI will be discussed in Sect. 3.3.5. 6. Target Detection and Tracking. The ability to do so is crucial to molecular communication networks. Target tracking aims to detect and track targets as they move, whereas target detection is a capability of molecular communication systems to detect a target in a particular environment. Targets in nanomedical applications may include disease locations, infections, infectious microbes, or biochemical weapons that pose a hazard to the environment; prompt target identification and monitoring are crucial for providing rapid therapies or conducting additional environmental investigation [4, 49, 51]. Even though it is one of the most promising fields of science, MCs pose a difficult integration problem to create a strong and stable communication system. Learning from biological systems is an effective strategy; for instance, self-organization, feedback mechanisms, modular architecture, evolution and adaptation, as well as other systems biology-related phenomena may be used to design and create a robust system. Moreover, some of the information and tools from communication engineering may be used to develop communication

3.3 Molecular Communication Systems

81

protocols or tools. Determining a resilient design architecture that enables communication components to self-organize and function despite the influence of noise and creating protocols and interface mechanisms are therefore general research issues in this field [45].

3.3.2 Theoretical Modeling of Molecular Communication 3.3.2.1

Random Walk

The quality of molecular communication can be measured, contrasted, and improved using theoretical models, such as average latency (i.e., average propagation delay), jitter (i.e., variation in latency), and loss rate (i.e., the probability that a molecule transmitted by a sender bio-nanomachine is not received by the intended receiver bio-nanomachine.) New molecular communication protocols and mechanisms can be created using theoretical models. In order to progress the field of molecular communication, the field of biophysics has produced a number of theoretical models (such as quantum-, molecule-, cellular-, and organ-level dynamics models) and computational techniques [49]. The single mechanism used in the first class of molecular communication is a random walk (i.e., no directional drift of information molecules and no chemical reaction of information molecules during propagation). The most fundamental mechanism for molecule propagation in molecular communication is the random walk. A molecule can spread using the random walk without the need for any extra mechanisms. Biology contains several instances of this category of molecular communication. One illustration is the spread of intracellular metabolites between cells. Another illustration is the spread of DNA-binding molecules (like repressors) over a DNA segment in search of a binding site. For simplicity, let us start with a one-way molecular communication in a semi-infinite interval .(−∞, d] with a sender bio-nanomachine at .x = 0 and a receiver bio-nanomachine at .x = d(> 0). The sender bio-nanomachine encodes information onto an information molecule and releases it at time .t = 0; the molecule then randomly walks in the environment; and upon arrival of the molecule, the receiver bio-nanomachine decodes the information from the molecule [49]. It is possible to define the latency in this class of molecular communication as the moment the molecule initially contacts the receiving bio-nanomachine, known as first passage time (FPT) [52], and its probability density function is given by Nakano et al. [49]:  f (x) =

.

0 √ d exp 4π Dt 3



−

d2 4Dt



(t = 0) (t > 0)

(3.1)

where D is the diffusion coefficient of the molecule. Figure 3.8 illustrates the probability density function of the latency for various distances d between sender

82

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.8 The probability density function of the latency in a semi-infinite interval .−∞, d for various sender and receiver bio-nanomachine distances .d = 1, 2, 4, 8(μm), D = 0, 1(μm2 /s)

and receiver bio-nanomachines. Figure 3.8 demonstrates that the latency is greatly affected by the distance d. The  ∞ average latency for a receiver bio-nanomachine at any location is infinity (i.e., . 0 tf (t)dt = ∞), which indicates that a receiver bionanomachine is expected to wait for an infinitely long amount of time to receive the molecule [61]. The jitter is also infinite. The loss rate can be obtained from T .1 − 0 f (t)dt, assuming that the receiver bio-nanomachine waits for the time duration T [49]. The quality of molecular communication differs significantly when the environment is bounded. Figure 3.9 shows the probability mass function .p(t) of the latency .(t) obtained from numerical simulation in a finite interval .[0, d] for various d where .x = 0 is reflecting. The average latency in this case is finite and is equal ∞ (i − (d 2 /2D))2 p(i), where . is to .d 2 /2D. The jitter can be calculated as .i=0 T p(i) under the the simulation time step length. The loss rate is given as .1 − i=0 same assumption T [49].

3.3 Molecular Communication Systems

83

Fig. 3.9 The probability mass function of the latency in a finite interval .[0, d] for various .d = {1, 5, 9} .(μm), D = 0, 1(μm2 /s)

3.3.2.2

Random Walk with Drift

A random walk with drift is the second category of molecular communication we investigate. Information molecules may experience a directional drift that propagates molecules continually in the drifting direction. Our bodies are a prime example of this category of molecular communication. Hormonal chemicals are

84

3 Analysis of the Molecular Physical Layer’s Tasks

released by body cells, travel via the bloodstream, and spread to distant target cells dispersed throughout the body. This category of molecular communication is also the active mode of molecular communication, in which motor proteins carry and propagate molecules in a specific route from a sender bio-nanomachine to a receiver bio-nanomachine [49]. When a molecule propagates in a fluid medium modeled as a semi-finite interval 2 .(−∞, d], the probability density function of the latency is given by replacing .d in 2 Eq. 3.1 with .(d–vt) , where .v(≥0) is the velocity of the fluid medium. Figure 3.10 shows the probability density function of the latency for various .v > 0 and demonstrates that the fluid medium becomes an effective medium to propagate molecules over long distances. For the fluid velocity .v > 0, the average latency is .d/v, and the expected latency decreases in proportion to the inverse of v. The jitter is .Dd/2v 3 and diminishes quickly as v increases. The loss rate can be obtained using Eq. 3.1 where .d 2 is replaced with .(d–vt)2 like the semi-finite interval case in Sect. 3.3.2.1 [49]. The physical concept of advection, which refers to molecular transport aided by bulk movement of the entire fluid, including the molecules of interest, is vital to note when it is mentioned to MCs with drift. Endocrine signaling in blood arteries and the control of fluids in microfluidic channels are two examples [49, 62, 63].

Fig. 3.10 The probability density function of the latency in a fluid medium for various fluid velocity .v = {0, 0.1, 0.2, 0.4} (μm/s), D = 0, 1(μm2 /s), and .d = 4(μm)

3.3 Molecular Communication Systems

3.3.2.3

85

Random Walk with Reaction by Amplifiers

A random walk combined with chemical reactions via amplifiers makes up the third class of molecular communication. By boosting the quantity of informationpropagating molecules, environmental amplifiers can improve the accuracy of molecular propagation. The environment contains amplifiers, which interact with molecules spreading there. Amplifiers create a replica of the molecule as a result, and it spreads across the environment. Using protein molecules like those in charge of amplifying calcium ions, adenosine triphosphate (ATP), and cyclic adenosine monophosphate may be able to facilitate this form of molecular communication (cAMP) [49, 62, 63]. To see the impact of amplifiers on the quality of molecular communication, consider a finite interval .[0, d] with a sender bio-nanomachine at .x = 0 and a receiver bio-nanomachine at .x = d(x = 0 is reflecting.). Repeaters are placed uniformly over the interval where the inter-repeater distance is l. Each repeater is activated when a molecule arrives at the repeater; once activated, it releases N molecules. (The activation is only one time for each repeater.) The probability mass functions .p(t) of the latency .(t) obtained from numerical simulation for various l and N are given in Figs. 3.11 and 3.12, respectively. Figure 3.11 shows that the latency decreases as more amplifiers are placed. Figure 3.12 shows that the latency also decreases as a larger number of molecules are released by an amplifier. The

Fig. 3.11 The probability mass function of the latency for various inter-repeater distances .l = {d/2, d/4, d/8} at .N = 2, d = 10(μm) and .D = 0, 1(μm2 /s)

86

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.12 The probability mass function of the latency for various numbers of molecules that a repeater releases .N = {2, 5, 10} at l = d/4, d = 10(μm), and D = 0.1(μm2 /s) ∞ ip(i), and the jitter and loss rate can be obtained from average latency is .i=0 .p(t) like the finite interval case in Sect. 3.3.2.1 [49, 62, 63].

3.3.3 ISI, Noise Analysis, and Error Control Techniques Based on [6, 64], the following paragraphs describe some general techniques briefly (avoiding specific situations that may not be applicable to all MC systems) to overcome these errors [65]:

3.3.3.1

Molecular Communications and the Impact of Noises

Given that stochastic chemical reactions control it, molecular communication noise must be controlled using a characterization of environmental dynamics. The primary source of mistake is due to effects brought on by the stochastic behavior present in chemical reactions. Many chemicals, proteins, amino acids, and other substances

3.3 Molecular Communication Systems

87

interact with one another. The amount of active cells in microscopic tissue volumes influences the dynamic behavior [6, 64]. Due to computing restrictions, a lack of control at that scale, and their reliance on specific molecule types and communication systems, nanomachines are unable to execute complicated coding schemes. Information coding techniques fall into two categories. The first is based on the type of particle used (such as neurotransmitters, intracellular messengers, or DNA molecules), while the second is based on how they use the particles as a means of propagation (e.g., free diffusion and cell-cell signaling). The tissue signaling environment is not addressed by the encoding processes outlined above. Error control codes, interference mitigation, error detection, and error avoidance are examples of existing techniques. Each method depends on the employment of particles and propagation to encrypt information [6, 64]. MC has its difficulties, such as the anticipated computing limitations. The chance of a communication error is increased by the high route loss rate and loud transmissions. Hence, creating a trustworthy MC is difficult. Diffusionbased MC enfolds distinct techniques and cellular signaling types, such as free diffusion communication and .Ca 2+ signaling. Free diffusion follows the behavior of molecules suspended in a fluid that moves in the absence of chemical reactions. 2+ signaling lies in a cell-to-cell communication usually mediated by gap .Ca junctions after a chemical reaction (i.e., reaction-diffusion). Due to more or fewer information carriers, different signaling types may involve varying numbers of chemical events, giving each signaling scheme its own propagation characteristics [6, 64].

3.3.3.2

A Classification of Noises and Their Sources

The most common sounds in MCs (explained in detail in this section) are categorized in Fig. 3.13, along with the primary methods for handling them and the mistakes they cause (detailed later). There are two viewpoints on noise in MCs: noise related to diffusion and noise related to signaling [6, 64]. 1. Free Diffusion-Related Noises. One of the primary sources of noise in free diffusion-based MC is the randomness from the movement of the information particles (Fig. 3.14). These

Fig. 3.13 Classification of error control: types of noises and techniques

88

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.14 Representation of a free diffusion-based MC system and noises

noises give rise to ISI, which has a significant impact on communication dependability. ISI is caused by the delayed arrival of information particles or by arrivals that are out of order, which causes transposition among symbols (also known as transposition noise, where symbols exchange positions rather than being distorted), making it difficult to detect and decode the signals that are being received [6, 64]. The transmitter-receiver distance is another characteristic that affects communication. Depending on the surroundings, the nanomachine’s position could change often (e.g., the bloodstream). Knowing the distance makes it possible to calculate the likelihood of eliminating noise and improving data transmission (e.g., frequency and concentration). Particles delivered and lost during the diffusion process before reaching the target constitute an erasing noise. There are a number of ways that molecules can be eliminated, such as by being absorbed by another receiver or becoming caught in biological structures [6, 64]. Unwanted interactions between information molecules and other molecules in the environment also result in noise (e.g., collisions breaking particles or generating new molecules). The information molecules’ structure is altered by these reactions, rendering the sent message undetectable to the receiver (decomposition noise). By using chemicals that have a comparable affinity to the receptors but are not used for communication, background noise or biological interference is produced [6, 64]. 2. Cellular Signaling-Related Noises. Molecular emission for cellular communication (e.g., .Ca 2+ signaling) promotes noisy conditions that produce data transmission errors. Cells serve as both sending and receiving nanomachines in cellular signaling networks (synthetic, hybrid, or natural). Synthetic biology offers methods for reprogramming the genetic code of cells to alter their composition and capabilities. Depending on the type of cell, communication gates may be used to govern the molecular movement of molecules from one cell to another. As a result, the properties of the cells and these gates are the focus of the noise. Gates that open and close between

3.3 Molecular Communication Systems

89

Fig. 3.15 Representing a cellular signaling-based MC system and noises

the cytoplasm of two adjacent cells are known as gap junctions. Through these gates, each cell is connected to a number of other cells, and because of their stochastic behavior, diffusion often travels to cells other than the destination cells, causing losses (erasure noise). When a gate unexpectedly closes during transmission, in big cells where the quantity of space has an impact on the molecule’s speed of propagation, or with symbol transposition, ISI occurs in this communication through delayed molecules [6, 64]. Due to changes in molecule concentrations during intercellular (i.e., betweencell) signaling, molecular signaling may produce too much noise (Fig. 3.15). Both internal and external noises may occur during cell-cell communication. Due to the stochastic occurrences of chemical processes, there are ongoing changes in molecule concentrations within cells. Intercellular channels’ permeability causes external noise, which is caused by changes in molecular concentration originating from nearby cells [6, 64].

3.3.3.3

Error Control Techniques

Traditional systems employ sophisticated error control strategies to guarantee accurate communication. These methods cannot be directly used for nanonetworks due to the unique restrictions placed on MCs. This section outlines various MC strategies for identifying, minimizing, and fixing communication channel faults. It classifies these strategies in a manner akin to four (Fig. 3.10), including [6, 64, 65]: • • • •

ISI mitigation Error prevention techniques Error detection codes Error control codes

90

3.3.3.3.1

3 Analysis of the Molecular Physical Layer’s Tasks

Intersymbol Interference Mitigation

The goal of ISI mitigation is to lessen the effects of diffusion’s stochastic character in order to boost channel reliability. Longer symbol durations or lower-energy symbols result in lowered molecule concentration in solutions. Low-complexity modulation and encoding methods are used to reduce ISI. Using modulation techniques based on particle release is one example. Three basic strategies lead to the lowering of the transposition error information [6, 64]: • Counting the number of released particles or alternating two different particles in two subsequent time slots • Sending the information in two types of molecule, one for symbol 1 and another for symbol 0 • Sending two different molecules simultaneously, subtracting their concentration, and identifying the equalized signal These methods demand that the receiver makes accurate distinctions between several symbols. For free diffusion systems, there are also ISI-resilient coding schemes. This method reduces particle arrival delays based on the quantity of molecules received. Yet, there is a pause between adjacent symbols. These mitigation strategies call for either more resources to manage the release or equipment capable of simple functions such as molecule storage or manufacturing, sensing, and accounting. These nanomachines can be created by genetically modified cells using synthetic biology, where specialized functionalities can be inserted into the cells, or artificial cells that blend biological and synthetic structural elements [6, 64, 66, 67].

3.3.3.3.2

Error Prevention Techniques

The different MC channels require distinct approaches to error prevention. The channel estimations’ expressive characteristic is called channel impulse response (CIR). The anticipated number of molecules in the receiver is used by MC systems to analyze CIR. Via CIR’s route loss, the receiver calculates the transmitter distance. The transmitter optimizes transmission parameters, such as longer pulses or greater signal amplitude for longer distances, to decrease erasing noise effects, so knowing this distance is crucial to lowering the channel error probability. Because nanomachines are not stationary in their positions, CIR in MCs fluctuates over time, making accurate CIR assessment challenging and pointless. This method also potentially uses many transmissions and takes up a lot of channel space. Measuring molecule propagation time using the propagation delay metric is another method for calculating the distance. When a feedback signal is requested, the receiver waits for a response before estimating the distance using the propagation delay. This method requires the transmitter to have more processing power [6, 64, 66, 67].

3.3 Molecular Communication Systems

3.3.3.3.3

91

Error Detection Codes

Parity check has been used in proposals for free diffusion-based systems to identify errors. Sending concentrations of two different molecule kinds using a modulated parity check encoder is one method. In this, the data is encoded using a threesymbol codeword that represents a binary code (i.e., the two states of the cell). Its decoder calculates the information’s a posteriori log-likelihood ratio and uses symbols to look for potential errors. In order to create a cell with branching genetic circuits—a system of interconnected chemical events involving genes and molecular species—to encode and regulate codeword symbols, this proposal draws on ideas from synthetic biology [6, 64, 66, 67]. Similar to how the Hamming code exhibits minimal complexity, MCs benefit from this. However, there is an energy cost when used with several random number generators and for encoding and decoding procedures. For MCs, a few Hamming code derivations have been proposed. The Hamming distance, for instance, creates channel codes based on the variety of symbols present in two binary sequences. In MC channels without concentration coding, the Hamming distance and channel timing-based decoding successfully combat additive noise sources. It is inadequate to deal with erasure errors, though. The Hamming weight codes calculate the distances between nodes and use the string’s weight to find and minimize transposition errors. The Hamming code (LT Hamming) and LT erasure codes prevent information loss in communication contexts of molecular scarcity (e.g., diffusive channels with erasure noises). These methods need numerous random number generators, which use energy and resources. Error detection is permitted, but not error correction [6, 64, 66, 67].

3.3.3.3.4

Error Correction Codes

Bit error rate reduction is the goal of error-correcting codes. Backward error correction (BEC) and forward error correction (FEC) are the two methods used for this purpose. The FEC frequently uses error control coding (e.g., Hamming and Reed-Solomon codes). It is crucial to take into account computational complexity and energy-efficient communication due to the projected constraints of nanoscale devices. By lowering the average weight of codewords in diffusion-based MC, minimum energy codes (MECs) with Hamming distance constraints can reduce energy usage. The average codeword weight energy increases as the average code weight lowers. Because transmitting symbol 0 costs no energy, using MECs in combination with on-off keying (OOK)2 modulation—transmitting molecules only during symbol 1 periods and not at all during symbol 0 periods—is appropriate for diffusive systems. MECs as channel codes enhance performance but necessitate lengthy codewords. In terms of the bit error rate produced by ISI, they perform better

2 This

type of modulation will be explained in Sect. 3.3.5.

92

3 Analysis of the Molecular Physical Layer’s Tasks

than Hamming codes. To enable applications, the complexity of coding systems must be kept low. Codes based on linear factors, such as the Reed-Solomon code, are among the most effective methods for error correction, although they consume a lot of energy [6, 64, 66, 67]. For MC, BEC approaches like stop-and-wait automatic repeat requests are available (SW-ARQ). In certain ideas, the SW-ARQ approach is used in free diffusion systems while taking into account the memory and computation limitations of nanomachines. To increase dependability, variations of this approach deliver repeated messages and acknowledgment (ACK). According to the method, the ACK control message carries out a number of functions, including causing the transmitter to transmit the next piece of information, pausing and releasing information molecules, and altering the kind of molecule. Feedback-based systems, like SW-ARQ, regulate the flow and guarantee that the receiver receives a specific amount of information particles while prohibiting the transmitter from releasing molecules more quickly than the receiver can respond. This prevents molecules from staying in the channel, crashing into one another, or diffusing away and resulting in loss or ISI, but it results in more time and energy being used. Each technique has drawbacks and benefits, and the choice must take the application into account [6, 64, 66, 67].

3.3.3.3.5

Error Prevention Technique for a Cell Signaling-Based MC System

Promising methods for engineering data transfer in MCs have been influenced by the natural signaling process of human body cells. Calcium signaling-based MCs are one illustration. Yet, establishing dependable communication between nanomachines in a noisy stochastic channel presents numerous difficulties. Lowcomplexity error control techniques are acceptable for channels based on cell signaling due to nanomachine restrictions [6, 64, 66, 67]. In light of cell signaling systems, this section suggests using several molecule types as information carriers to encode data (i.e., juxtacrine signaling). It takes advantage of the diversity of molecules and their relationships to increase reliability in MCs because natural cells display a wide range of chemicals and their relevant pathways. The proposal lies in encoding information using multiple information carriers, such as inositol triphosphate .(I P3 ), calcium .(Ca 2+ ), and potassium + .(K ) molecules (i.e., molecules commonly found in natural cells, e.g., astrocytes, epithelial, and smooth muscle, among others). The dominant alternative to lower mistake probability and produce low-complexity signaling MCs is natural molecular diversity and their nonlinear interactions, according to the research premise. The selection of chemicals is bio-inspiring and is the result of extensive research by MCs on the underlying biological processes that may be impacted. As they are already present in the human body, these particular molecules have not before been used for error control and are biocompatible and practical for intra-body artificial MCs [6, 64, 66, 67].

3.3 Molecular Communication Systems

93

The binary modulation technique uses concentrations of two types of information molecules, .I P3 to represent symbol 0 and .Ca 2+ to represent symbol 1, to reduce ISI effects and errors in communication. Positive values for each molecule depend on a concentration threshold, and any value above it is considered to be a symbol 1 2+ ) or symbol .0 (I P ), according to the molecule whose concentration was .(Ca 3 measured. Using different molecules to represent symbols 1 and 0 reduces the probability of error caused by noise in the channel [6, 64, 66, 67]. The use of many carriers improves the receiver’s ability to distinguish between two separate symbols. Errors can happen more quickly when modulation only employs one type of molecule and represents the symbol 0 by silence (i.e., by not transmitting molecules). This is because the noise makes it difficult to identify the symbol 0 and makes it difficult to decode the received signals. T x decides which molecule to send depending on the previously sent transmission symbol and current symbol. However, it must satisfy the size of the codeword length .l ≤ 7 symbols (i.e., .l = 3 message symbols, .l ≤ 3 parity check symbols, and .l = 1 symbol for a start signal). A time slot .(t) is reserved for the release of each symbol (i.e., symbol time interval) based on the propagation time. The natural mechanism of the cells causes the molecules to arrive through the intercellular channels to be quickly expelled from the cells; then, even if the time for release is added, communication tends to be fast [6, 64, 66, 67]. Given the asynchronous communication and the difficulty of feedback signals, the proposed scheme considers a single symbol of .K + at the beginning of the codeword to indicate a new transmission. Error detection is performed using a parity check scheme. The parity symbols are added in the last positions of a string of the binary code and comprise a concentration of .K + . The parity symbol ensures that the total number of symbol 1 in the string is even if three consecutive .K + symbols are included at the end of the codeword or odd if only one symbol is added (Fig. 3.16). For example, to transmit a three-symbol message 101, the transmitter sends the 2+ molecules to symbol 1 and then releases .I P to symbol 0 and .Ca 2+ for the .Ca 3 next symbol. Here, parity adds three .K + symbols at the end of the codeword to indicate that the total number of ones (i.e., .Ca 2+ ) in the string is even. The receiver nanomachine recognizes that when .I P3 or .Ca 2+ are followed by .K + , it indicates the parity symbol. Once all messages begin with .K + and there are no other cases where .I P3 or .Ca 2+ are followed by .K + , the parity symbol also indicates the end of the message. The same message is repeated at least three times with a long period of the interval .Tb between each repetition. Thus, the receiver checks parity and discards error messages [6, 64, 66, 67]. Based on [68], the proposed error prevention technique has been studied through mathematical simulations for a cellular signaling-based MC system. It is assumed in [6, 64, 66, 67] an asynchronous single-hop MC system is composed of a hybrid transmitter and receiver nanomachine (i.e., a biological nanodevice with electronic parts) and channel. Rx measures the concentration of molecules upon contact and checks the parity symbol. Cell-to-cell communication occurs when the gap junction in the cells opens and .I P3 , .Ca 2+ , and .K + molecules propagate through the cytosol. In [6, 64, 66, 67], intracellular noises have been modeled for .I P3 and .Ca 2+ based

94

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.16 Error prevention technique

Fig. 3.17 Capacity and path loss

on [69]. Figure 3.17 compares the proposed technique with OOK modulation to transmit both molecules in symbol 1 periods and not transmit molecules in symbol 0. The comparison follows a perspective of information theory. In MCs, molecules may not arrive at the receiver due to their diffusion direction probability in intercellular channels. Thus, path loss was applied to analyze this behavior using different molecules as information carriers, considering the intracellular/intercellular channel

3.3 Molecular Communication Systems

95

noises. Shannon’s entropy quantifies the communication capacity in terms of symbols considering a discrete memoryless channel [6, 64, 66, 67]. In [6, 64, 66, 67], the state transition probabilities for Rx and T x are defined. For T x, they consider the release of .Ca 2+ or .I P3 molecules. For the receiver, in [6, 64, 66, 67], another two states are considered to represent when the amount of received molecules (i.e., .Ca 2+ /I P3 ) changes cell state, transitioning from symbol 1 to symbol 0. The amount of shared information is measured by mutual information. Findings show that when information encoding employs several information carriers in conjunction with the suggested error avoidance strategy, end-to-end capacity improves by up to 10%. The disruption of communication is brought on by intracellular noise. The error avoidance method, which hasn’t been extensively adopted, merely lowers path loss by establishing a dual signaling mechanism employing several carrier molecules. The findings pave the way for more thorough investigations into the number of molecules that can be interlaced to maximize communication performance in MCs [6, 64, 66, 67].

3.3.4 Analysis of Physical Estimation Parameters Due to the significance of numerous crucial applications, particularly in the field of medicine, the assessment of physical parameters in molecular communication is a crucial topic. Hence, in [70], the MC systems can be deployed in a blood vessel environment for healthcare applications. In this case, measuring blood flow velocity can assist in taking blood pressure readings. Determining the blood composition and spotting significant variations in blood cell counts can also be done with the aid of determining diffusion coefficients for substances. Additionally, because molecule degradation differs depending on pH level, calculating the rate of molecule breakdown can aid in determining the pH level of blood. The estimated parameters include the number of molecules released, the rate at which they degrade, the flow rate, the diffusion coefficient of molecules, the time at which they release themselves, the offset between the clocks of the transmitter and the receiver, the beginning of each symbol interval, and the signal-to-noise ratio (SNR). Unlike parameter estimation, channel estimation focuses on estimating MC systems’ channel impulse response (CIR). Here, the CIR is defined as the probability of observing one molecule at the receiver at time t when molecules are impulsively released at time .t0 = 0. The CIR is important for the design of equalization and detection schemes in MC systems [70]. Most studies considered estimation schemes in molecular communications via diffusion MCvD system as shown in [70]. This system is explained as shown in Fig. 3.18, where one point T x communicates with one spherical Rx with radius .rR in an unbounded three-dimensional (3D) environment. The Rx center is d away from the T x. The T x is considered a point source that releases molecules into the environment [70].

96

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.18 Illustration of the MCvD environment where one point T x communicates with one spherical Rx in a three-dimensional environment

• Propagation Channel Modeling In [70], the propagation channel modeling is assumed outside the T x, and the Rx is filled with a fluid medium. Once molecules are released, they diffuse randomly in the propagation environment. The movement speed of molecules is determined by the diffusion coefficient, denoted by D, which is affected by the fluid medium’s temperature, the fluid’s dynamic viscosity, and Stoke’s radius of molecules. In [70], it is also assumed that the fluid medium has uniform temperature and viscosity such that D can be modeled as a constant value. It is noted that a more complex MC environment can incorporate flow with a constant velocity v and degradation of molecules, i.e., molecules of type A can degrade into some other molecular species  with a constant degradation rate k.  cannot be recognized by Rx. • Receiver Modeling In [70], for the receiver modeling, it is denoted .h(t) as the CIR of the end-toend channel, and they use three models for Rx modeling: * The first one defines a transparent Rx modeling that does not impede the diffusion of molecules nor interact with molecules. It is counted as the number of free molecules within the Rx volume as the received signal. The CIR of the transparent Rx is given by Equation [70]:  h(t) =

.

VRX exp (4π Dt)3/2

−

d2 4Dt

 (3.2)

3.3 Molecular Communication Systems

97

where .VRX is the volume of the Rx and .VRX = 4/3π rR3 for the spherical Rx. It is noted that .h(t) incorporates the flow and degradation of molecules, and it is also noted that Eq. 3.2 is accurate when the Rx is sufficiently far away from the T x, i.e., d is very large relative to .rR . Thus, it is reasonable to assume that the concentration of molecules at every point within the Rx equals the concentration at the central point of the Rx. If the Rx is close to the T x, the uniform concentration assumption does not hold. In this case, .h(t) is given by Eq. 3.3 [70]: h(t) = .

1 2

       √ rR −d rR +d Dt √ √ √ erf + erf + r π x exp − (4Dt) (4Dt) R   2 (rR +d) +exp − 4Dt

(rR −d)2 4Dt



(3.3) where erf .(·) denotes the error function. In the previous equations, it is noted that Eq. 3.2 is more widely applied than Eq. 3.3 due to its simplicity and Eq. 3.2 provides an accurate approximation for Eq. 3.3 if .rR < 0.15 d [70]. ¯ .(t) as the expected number of molecules observed Now, it is denoted N within the Rx volume at time t. If an impulse of .Ntx molecules is released ¯ .(t) is given by N ¯ .(t) = Ntx h(t). In Fig. 3.19, from the T x at time .t0 = 0, N ¯ .(t) versus time t by adopting Equation 3.2 is plotted. In that figure, a peak N number of molecules are observed within the Rx. In [70], it is denoted the time for reaching the peak number of molecules observed as .tmax . By taking the derivative of Eq. 3.2 with respect to t, .tmax calculated as .tmax = d 2 /6 D, then if this mathematical expression is substituted in Eq. 3.2, the mathematical √ 2π e −3 ¯ .(t)max is the expression to get the peak CIR is .hmax = (d 3 ) VRX . If N ¯ .(tmax ) = expected peak number of molecules observed within the Rx, then N Ntx hmax [70]. * In the second case of analysis in [70], the fully absorbing Rx modeling is presented; due to that in biological systems, many practical Rx surfaces may interact with the molecules of interest, e.g., by providing binding sites for absorption or other reactions. Hence, the transparent Rx model is oversimplified. One practical Rx model is the fully absorbing Rx. In this model, the Rx absorbs molecules as soon as they hit the surface. The fully absorbing Rx counts the total number of molecules absorbed as the received d−rR signal, where the CIR is given by .h(t) = rdR erf c( √ ). 4Dt * The third model in [71] is presented as reactive Rx modeling; this is the case where molecules that reach the Rx may participate in a reversible bimolecular second-order reaction with receptors over the Rx membrane. This type of Rx is named the reactive Rx, whose received signal is the number of activated receptors. The CIR of the reactive Rx is given by Eq. 3.3, which is proportioned by a long and very complex study in [71].

98

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.19 The number of observed molecules within the transparent Rx at time t versus time t, where .Ntx = 105 , rR = 0.5 μm, d = 4 μm, D = 1000 (μm2 /s)

• Noise Modeling As previously mentioned in Section 3.3.3 in [70], it is also mentioned that some factors affect the performance of estimation and treat these factors as noise during the estimation process. Hence, in the next paragraphs, this negative factor for molecular communication systems is analyzed from three perspectives: * The first one is the statistical distribution of the received signal. Thus, due to the random diffusion (RD) of molecules, the number of molecules observed at the Rx is a random variable (RV). This randomness influences the performance of estimation. Due to the independent diffusion of molecules, any given molecule released by the T x is observed by the Rx with a probability of .h(t). A binary state model applies, and the number of molecules observed at time t, denoted by .Nob (t), follows a binomial distribution with .Ntx trials and success probability .h(t). This is mathematically expressed as .Nob (t) ∼ B(Ntx , h(t)), where .B(N, p) represents a binomial distribution. Unfortunately, the binomial distribution is awkward to work with in MC systems. Therefore, current studies usually approximate binomial distribution as two distributions for the sake of mathematical tractability, described as follows [70]:

3.3 Molecular Communication Systems

99

(a) Poisson distribution. When the number of trials .Nt x large and the success probability .h(t) is small, .Nob (t) can be approximated as a Poisson RV, given by .Nob (t) ∼ P(Ntx h(t)) where .P(ϕ) represents the Poisson distribution with the mean of .ϕ. Based on this last mathematical expression, the probability mass function (PMF) of the Poisson RV (Ntx h(t))ξ exp(−Ntx h(t)) , where .Nob (t) is written as .Pr (Nob (t) = ξ ) = ξ! .P r(·) stands for the probability [70]. (b) Gaussian distribution. If the expected number of molecules observed, i.e., .N¯ ( t), is sufficiently large, the central limit theorem can be applied and approximate .Nob (t) as a Gaussian RV, given by .Nob (t) ∼ N (Ntx h(t), Ntx h(t)(1 − h(t))). The probability density function (PDF) of .Nob (t)is given by Equation (3) [70]:  .

(ξ −Ntx h(t))2 2Ntx h(t)(1−h(t))



exp − Pr (Nob (t) = ξ ) = √ 2π Ntx h(t)(1 − h(t))

(3.4)

* The second part of the noise analysis in [70] refers to external additive noise; in MC systems, the intended T x may not be the only source of molecules. Other sources, referred to as external sources, may also release the same type of molecules that influences the observation at the Rx and affects the estimation performance. Some examples of external sources include as follows [70]: (a) Multiuser interference. Noisy molecules are emitted by T xs in other MC systems. (b) Unintended leakage. Molecules can be leaked from membrane-bound containers, e.g., vesicles, within transceivers. A rupture can result in a steady or sudden release of molecules. (c) Output from unrelated biochemical processes. Biocompatibility of the MC system may require the selection of naturally occurring molecules. Therefore, other processes that produce the same type of molecules are noisy sources for the considered MC system. (d) Unintended reception of other molecules. Molecules that are highly similar to intended molecules may be recognized by the Rx. .Nsig (t) is the intended observed molecules and .n(t) as the observed noise molecules. Since intended molecules and noise molecules are indistinguishable, the total number of molecules observed at the Rx at time t is given by .Nobt (t) = Nsig (t) + n(t) [70]. The analysis of the statistical distribution of .n(t) is built upon the following assumptions: First, .n¯ is the expected number of noise molecules observed within the Rx. In [70], it is assumed that .n¯ is constant over the entire observation time.

100

3 Analysis of the Molecular Physical Layer’s Tasks

Second, the observation of one noise molecule at the Rx is independent of observations of other noise molecules. Third, the uniform concentration assumption holds for noise molecules at the Rx. These assumptions let modeling the number of observed noise molecules as a Poisson RV, due to the law of rare events (LRE), i.e., .n(t) ∼ P(n). ¯ * The third part of the analysis of noise in [70] is established as ISI and ILI analysis. ISI exists when the T x transmits multiple symbols to the Rx. Due to the RD of molecules, the molecules from previously sent symbols may arrive at the Rx in the current symbol interval, which influences the estimation in the current symbol interval. ILI exists for the multiple-input multiple-output (MIMO) MC system, where one .T x −Rx channel is influenced by molecules released from other channels. • Distance Estimation In this section, [70] classifies the current studies about distance estimation into two-way estimation and one-way estimation as follows: * Two-way estimation. To examine these schemes, estimate the distance between two transceivers labeled as T and R, respectively. Specifically, T first releases an impulse of molecules of type A at time .t0 . The diffusion coefficient of molecules A is .DA . When R detects the molecules A at time .t1 , it immediately transmits a feedback signal of an impulse of molecules B whose diffusion coefficient is .DB . At time .t2 , T detects the molecules B. T and R are regarded as transparent Rxs when they detect molecules B and A, respectively [70]: (a) Round trip time (RTT) protocols In RTT protocols, T measures the RTT, the sum of time required for the transmission from R to T . The first estimation scheme is the RTT protocol from peak concentration (RTT-P). In RTT-P, T transmits at time .t0 , and R detects the peak concentration of molecules A from T at time .t1 . T detects the peak concentration of the feedback signal with type B molecules at time .t

2 . The distance d is estimated as a function of the 6DA DB ˆ RTT .t2 − t0 as .d = D (t2 − t0 ), where .dˆ is the estimated value of A +DB d [70]. The second estimation scheme was named the RTT protocol from threshold concentration (RTT-T). Different from RTT-P, RTT-T defines threshold concentrations .HA and .HB for R and T to detect the number A of molecules observed, respectively. It is assumed that T transmits .Ntx B number of molecules number of molecules A and R transmits .Ntx B. R records .t1 when the number of molecules observed reaches the ¯ 1 − t0 )| threshold concentration .HA , i.e., .N(t A = HA by D=DA ,Ntx =Ntx ¯ = Nobs (t). Similarly, T records .t2 when the number assuming .N(t) of molecules observed reaches the threshold concentration .HB , i.e., A B ¯ 2 − t1 )| .N(t D=DB ,Ntx =N B = HB . If .DA = DB , Ntx = Ntx , and tx

3.3 Molecular Communication Systems

101

ˆ H A = HB , d is estimated as a function of the RTT .t2 − t0 as .d = 2 (N A )2 VRx DA (t2 − t0 ) ln( 3 3 2 T x 3 ) [70].

.

8π DA HA (t2 −t0 )

(b) Signal attenuation protocol from peak concentration In signal attenuation protocol from peak concentration, T transmits type A molecules to R, and R measures the peak concentration, denoted A . The relationship between .N A by .Nob,m ob,m and d can be obtained by A ¯ replacing .Nmax with .Nob,m . Similarly, R transmits type B molecules, B . By assuming and T measures the peak concentration, denoted by .Nob,m

2 A 3 Ntx VRx 16 A B ˆ ˆ .Ntx = N ) [70]. ob,m , .d is obtained as .d = 2π e ( B Nob.m

* One-way estimation The majority of recent studies have concentrated primarily on one-way estimate or the gathering of distance information from the received signal [70]: (a) Peak-based estimation In [70], the studies that perform the estimation via measuring the peak received signal or the time for reaching the peak received signal are reviewed. First, the estimation based on the peak number of observed molecules denotes .Nob,m at the transparent Rx. The distance d is 3 VRx Ntx 13 estimated as .dˆ = ) . The estimation performance of this ( 2π e

Nob,m

method is affected by the RD of molecules since the instantaneous observation is used to approximate its expectation. Second, for the first time, the release of molecules from the T x as a rectangular pulse, where the CIR at the transparent Rx is denoted by .hrec (t). By taking the derivative of .hrec (t) concerning t, d can be estimated via measuring the time for reaching the peak concentration at the Rx as

6Dtmax (tmax −Te ) ˆ ln( tmzx ) where .Te is the emission duration [70]. .d = Te

tmax −Te

As .tmax is involved in the estimation of [70], a perfect synchronization between the T x and the Rx is required. In [70], a low-complexity scheme that does not require synchronization via adopting two types of molecules is also considered. In this scheme, the T x releases type A molecules at time .t0 , and the Rx records the time of peak concentration A . Similarly, the T x transmits type B of molecules A, denoted by .tmax molecules at .t1 , and the Rx records the time of peak concentration of B . The relation between .t A and d and .t B molecules B, denoted by .tmax max

max and d is obtained, with an estimation of d, as .dˆ = 6DA DB (tRx −tT x ) DA −DB

B −t A and .t where .tRx = tmax T x = t1 −t0 . It is necessary to mention max that .tRx and .tT x are based on the time measurements at the Rx and the T x, respectively. Therefore, synchronization is not required [70].

102

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.20 Estimation in a cylindrical diffusive MC environment by using ring-shaped Rxs

According to [70], if an estimation is defined considering a cylindrical MC environment whose surface is reflecting as shown in Fig. 3.20, thus a point T x releases molecules and two ring-shaped Rxs with a certain width ware located on the cylinder perimeter with a radius .rv that is the same as the cylinder’s radius. After the molecules are released, they diffuse randomly with a constant diffusion coefficient D and are subject to the Poiseuille flow; then, advection and diffusion can transport molecules. If it is considered a diffusion-domain movement and only focused on the movement in the x axis, i.e., the channel is approximated as a 1D environment, in this situation, the estimation is performed in two scenarios: The first scenario is that the emission time of molecules is known. By measuring the time of peak concentration, the distance can be estimated by a single Rx as √ −w+ w2 −8(wvm tmax −2(vm tmax )2 −4De tmax ) ˆ .d1 ≈ , where .De is the effective 4 diffusion coefficient and .vm is the average flow velocity. The second scenario is that the emission time is unknown. Thus, .tmax and .d1 are unknown. In this scenario, .d1 is estimated by using two Rxs, and these two parameters can be obtained at the second Rx. Thus, .d1 can be mathematically derived by jointly solving the previous mathematical relation [70]. (b) Maximum likelihood (ML) estimation θ that maximizes the joint observation The ML estimation is to find . likelihood, which usually achieves high accuracy but requires high computational complexity, and the perfect synchronization between the T x and the Rx, here .xm , is the mth data of the data set, as the number of molecules observed at the transparent Rx at time .tm when molecules are released at time .t0 . It is assumed that each observation is independent and follows a Poisson distribution. Thus, .p(x; θ ) is

3.3 Molecular Communication Systems

103 xm

(Ntx h(tm )) given by .p(x; θ ) = M exp(−Ntx h(tm )). Using this last m=1 xm ! mathematical expression, d is estimated by taking the partial derivative of .p(x; θ ) with respect to d and setting it equal to 0 [70]. (c) Nontransparent Rxs An estimation scheme by applying a fully absorbing Rx is investigated in [70]. The Rx performs estimation by counting the number of molecules absorbed within a time interval. Specifically, an algorithm was proposed when the T x and the Rx are unsynchronized; then, two optimization methods were used to improve estimation performance. The first method uses molecules with a large diffusion coefficient, and the second method increases the number of emitted molecules. Also, in [70], an estimation is studied in an environment with multiple fully absorbing Rxs. As one fully absorbing Rx would impact molecules absorbed by other fully absorbing Rxs, an accurate derivation for the number of molecules absorbed at each Rx is cumbersome. Besides, in [70], a curve fitting method to obtain the expression for the number of absorbed molecules is adopted. Specifically, the nonlinear least squares method for curve fitting is used to obtain the distance, where the Levenberg-Marquardt (LM) method is adopted.

• Synchronization-Related Parameters The clock offset describes a time difference between the T x and the Rx. In a nanonetwork system, different nanomachines work based on their clocks. Hence, estimating the clock offset is crucial to establish a reliable communication link between synchronized T x and Rx. Here, the clock offset between two transceivers is denoted by T and R, respectively, via proposing a two-way message exchange mechanism. In the . th round of the message exchange, T sends molecules at time .T(1, ) , and R receives the message at time .T(2, ) . R then sends a feedback signal at time .T(3, ) , and T receives the signal at time .T(4, ) . After .α-rounds α of message exchange, T obtains a set of time instants . T1, , T2, , T3, , T4, =1 . By assuming that the propagation delay follows an inverse Gaussian distribution and Gaussian distribution, the joint PDF of the molecular propagation delay for the .α-round message exchange can be obtained. Based on the PDF, the clock offset is estimated via ML estimation. After that, R can be synchronized to T . Based on the joint PDF of the molecular propagation delay for multiple transmissions from the T x, the ML estimation estimates the clock offset. Furthermore, if it is considered clock offset estimation when both T x and Rx are diffusive mobiles, then after molecules are released from the T x, the Rx counts the number of arrived molecules for M times. Hence, it estimates clock offset by using the least squares method that finds the clock offset to minimize the sum of differences between the mean of the CIR over the varying distance and M observations [70]. The clock offset estimation is adequate to achieve synchronization only if the clock offset is fixed. To overcome this issue, it is considered the estimation of the start time of the .ηth symbol interval, denoted by .ts [η]. Hence, the ML estimation scheme is proposed to estimate .ts [η]. Considering that each observation within

104

3 Analysis of the Molecular Physical Layer’s Tasks

this interval follows a Poisson distribution, .ts [η] is estimated based on the joint PDF of multiple observations. Due to the high complexity of the ML estimation, three suboptimal low-complexity estimation schemes are proposed. The first suboptimal estimation scheme is a linear filter-based scheme that finds .ts [η] to maximize the expected mean of multiple observations. The second is the peak observation-based scheme that estimates .ts [η] based on the peak observed at the Rx. The third is the threshold-trigger scheme that determines .ts [η] when the observation is larger than a predefined threshold. Notably, in [70] it is considered the impact of the external additive noise and ISI on these estimation schemes and also it is considered the transmission of molecules with a faster diffusion coefficient than information molecules to realize synchronization, where .ts [η] is estimated as the time when the peak concentration of faster molecules is detected at the Rx [70]. • Signal-to-Noise Ratio (SNR) An estimated SNR in the MC system is considered when the noise is induced by the ISI. Hence, the SNR was defined as .SNR = PPSn where .PS represents the power of the intended received signal at the transparent Rx and .Pn represents the power of noise due to the ISI. The power in MC can be interpreted as the square number of molecules. Thus, the SNR is a function of the number of emitted molecules .Ntx and the noise variance, denoted by .σn2 , where the noise is regarded as a Gaussian RV. According to the invariance property of the ML estimation, estimating a function with multiple unknown parameters is equivalent to estimating individual unknown parameters. Therefore, the SNR is estimated using the ML estimation of .Ntx and .σn2 n based on the joint PDF of the received signals at the Rx. In addition, the CRLB (Cramer-Rao lower bound, that is, a lower bound on the variance of any unbiased estimation scheme) was derived 2 , σ 2 ] [70]. when .θ = Ntx n • Channel Estimation Here, it presents the CIR estimation problem and then reviews different CIR estimation schemes. * Problem Formulation For this scheme, the MC system is shown in Fig. 3.18. At the beginning of each symbol interval, the T x releases .Ntx molecules if the transmitted symbol is “1” but does not release any molecule if the transmitted symbol is “0.” Taking into account the effect of ISI, it is assumed that the inputU c [q] + c [q], where output relationship of the MC system is .z[q] = u=1 u n .z[q] is the number of molecules detected at the Rx in symbol interval q, U is the number of memory taps of the channel, and .cu [q] is the number of molecules observed at the Rx in symbol interval u, due to the release of .b [q − u + 1]Ntx molecules by the T x in symbol interval .q − u + 1, where .b[q] [0, 1] is the transmitted symbol in symbol interval q. Therefore, .cu [q] can be well approximated by a Poisson RV with the mean of .c¯u b[q − u + 1]. Moreover, .cu [q] is the number of external additive noise molecules detected

3.3 Molecular Communication Systems

105

by the Rx in the symbol interval q. Let .b = [b[1], b[2], . . . , b[Q]]T be a training sequence of length Q [70]. For convenience of notation, it is defined .z = [z[U ], z[U + 1], . . . , z[Q]]T , .c¯ = [c¯1 , c¯2 , . . . , c¯U , c¯n ]T is the CIR of the channel, and .fz (z|c, ¯ b) is the PDF of the observation z conditioned on a given channel .c¯ and a given training sequence b. The goal of channel estimation is to estimate ¯ based on the vector of random observations z [70]. .c * Pilot-Based CIR Estimation A pilot-based CIR estimation scheme is studied in [70] where the transmission of a known training sequence of pilots is required for the estimation and calculation of the corresponding CRLB: (a) ML estimation The ML CIR estimation scheme aims to find the CIR that maximizes the likelihood of observation vector .z. In particular, the ML estimation is given by .cˆ¯ML = argmax fz (z|¯c b). It is assumed that .z[q] is a Poisson c¯ ≥0  ¯ T bq and RV with the mean of .z¯ [q] = c¯n + U u=1 c¯u b[q − u + 1] = c T .bq = [b[q], b[q − 1], . . . , b[q − U + 1], 1] . Under this assumption, Q

(¯cT b )z[q ]

q fz (z|¯c, b) is given by fz (z|¯c, b) = q=U z[q]! exp(−¯cT bq ) [70]. (b) Least sum of squared errors (LSSE) CIR estimation The least sum of squared errors CIR estimation scheme aims to choose .c¯ that minimizes the sum of the squared errors for the observation vector .z. Here, the error vector is defined as . = z − E {z} = z − B¯c where .B = [bU , bU +1 , . . . , bQ ]T . In particular, the least sum of squared argmax || ||2 [70]. errors CIR estimation can be written as .cˆ¯LSSE = c≥0 ¯ (c) CRLB With the estimation error vector defined as .e = c¯ − cˆ¯ , the classical bound CRLB for the deterministic .c¯ provides the following lower

−1  on 2 the sum of the expected square errors .E {e} ≥ tr I (¯ c ) =    T −1 b b q Q , where .tr {·} denotes the trace of a matrix [70]. tr q=U cT bq .

q

* Semi-blind CIR Estimation Here, the scenery establishes a transmission of Q pilot symbols followed by the transmission of D unknown data symbols, denoted by the vector .β = [β[1], [2], . . . , β[D]]T . Both data-carrying and pilot-carrying observations are incorporated into the estimation process in semi-blind CIR estimation techniques. Instead of including data-carrying observations, pilot-based estimation solely takes into account the received pilot-carrying observations. The estimation’s accuracy and/or data rate can be considerably improved by including data-carrying observations [70]. The data vector .β constitutes hidden information at the Rx. The EM estimating strategy alternates between obtaining the conditional expectation

106

3 Analysis of the Molecular Physical Layer’s Tasks

of the complete data log-likelihood and maximizing the result with regard to the desired parameters, starting with the initial guess. The EM estimation scheme’s th iteration is divided into two parts. The complete data loglikelihood function’s expectation is what is determined in the first step, called the expectation (E) step. The maximizing step (M step), which obtains an updated estimate of the CIR, comes next. In this step, the posterior probability of the hidden data is maximized. There are additional suggestions for two decision-directed (DD)-based semi-blind estimate systems. The goal of the DD strategy is to execute data detection using the channel estimate obtained from pilot-based estimation. For the purpose of doing another cycle of channel estimation, the detected data is treated as a fresh set of pilots. It is significant to note that the analytical derivation of the CRLB in a semi-blind estimate can be highly difficult due to the complexity of the log-likelihood function of observations when the statistics of data symbols are taken into consideration [70]. * Pilot-Based Estimation Versus Semi-blind Estimation According to the simulation results in Fig. 3.21, the semi-blind estimation schemes outperform the current pilot-based ML and the least sum of squared error estimation schemes in terms of mean squared error (MSE). Also, the semi-blind estimation schemes can substantially reduce the pilot overhead compared to the best-performing pilot-based estimation schemes by more than 60% for the case of EM and 55–44% for DD-based estimation. The EM-based semi-blind estimation scheme provides the highest estimation accuracy. The DD-based semi-blind estimation scheme performs almost midway between the pilot-based ML and the EM-based semi-blind estimation schemes but achieves a lower computational cost than the EM-based semi-blind estimation scheme [70]. Fig. 3.21 The MSE of different CIR estimation schemes versus pilot sequence length: LSSE, least squares; ML, maximum likelihood; P-CRLB, pilot-based CRLB; DD-ML, decision-directed ML; DD-LS, decision-directed LS; SB-CRLB, semi-blind CRLB

3.3 Molecular Communication Systems

107

3.3.5 Molecular Modulation Several strategies can be utilized to address the ISI brought on by the molecules’ slow passage through biological systems. There have already been initiatives to get rid of the ISI by changing the detection interval [4]. Depending on their qualities, these techniques can be grouped into three groups: the shift-.τ method, truncating the symbol duration in advance, and extracting a small portion of the symbol duration. First, the shift-.τ method means that the receiver can shift its absorbing/counting interval backward by .τ seconds to avoid the strong ISI region. Receivers can be designed with confidence using the approximate optimal reception delay of the absorbing receiver, which must be specified. Second, as proposed for the MCvD system made up of a transmitter, an absorbing receiver, and an interference source, truncating the symbol interval in advance means that the detection process will be stopped before the conclusion of a symbol period. The authors discovered a suitable detection interval that can successfully counteract interference from unwanted transmitters. Finally, because only a piece of the symbol duration is extracted, the detection procedure will begin and end later than expected [4]. Modulation-based strategies are used in other ISI defense mechanisms. As a result, the principles relating to molecular modulation will be examined in the paragraphs that follow: Due to the nature of the physical molecular link, it is necessary to have some mechanism to control the transmission of information. Thus, some works model transmitter [72, 73] and receiver [74–77], and most of the research has been focused on modeling the channel as well as the modulation and coding schemes necessary for setting up a reliable communication link [78–85]: • Transmitter models. Only a small number of papers have included MC transmitters in information and system theoretic models. This is because there are no models available for the transmitter. Future channels must take into account transmitters with faulty transmission and transmitters with restricted energy, particle, and resource availability [78]. • Receiver models. In the literature, very few works have considered the receiver as part of the system model. However, as some experimental results have demonstrated, sensors could significantly affect the system model [86]. More work is required on modeling different sensors and detectors for MC receivers [78]. • Channels with memory. The majority of practical MC channels have lengthy memory and ISI. To make the formulations and models manageable, memoryless channels without ISI have been assumed in several publications on information theoretic MC channels. Future research must take a look at modifying some of these presumptions [78, 87]. • General capacity of MC channel. A significant open topic is the generic memory capacity of the MC channel for all conceivable modulation schemes. Nevertheless, it should be recognized that this is a challenging issue to resolve [78, 87].

108

3 Analysis of the Molecular Physical Layer’s Tasks

• Capacity of MC channel with reactions. The information particles in some MC channels can interact with one another or with other particles in the channel. These effects have been taken into account in very few research when defining channel capacity expressions [78, 87]. • Optimal detection and coding. Although the capacity of some straightforward MC channels is known, in some circumstances, it is not evident how this capacity might be reached. There is more work to be done on the best detectors and coding for MC [78, 87]. MC channels have been studied in the literature and appraised using simulations or models [88]. The most difficult task is to use experiments to validate these modulation strategies. For instance, the majority of modulations make the unrealistic assumption that there are flawless transmitters and receivers. It is uncertain how much a defective transmitter or receiver affects the modulation scheme’s performance. Together with this overarching general direction, the following intriguing open issues should be addressed [78, 87]: • Synchronization. Several of the modulation techniques suggested for MCs would call for synchronization between the transmitter and the receiver. This issue has been addressed in several earlier publications, but more research is necessary [78, 87]. • Channel state estimation. Channel state information is needed at the transmitter or receiver for some particular modulation schemes. With MCs, channel state information is often an estimation of the diffusion coefficient or the distance between the transmitter and the receiver [78, 87]. • ISI cancellation. The performance of modulation is substantially impacted by the ISI found in many MC channels. Hence, to address this problem, ISI cancellation techniques must be used [78, 87]. • Higher-order modulation. It may yet be possible to create unique high-order modulation methods that can enhance performance even if a variety of modulation strategies have been presented for MCs [78, 87]. Different modulation techniques are proposed in the MC literature (classified according to the physical property used to encode bits into the transmitted symbols through information molecules in a diffusion molecular environment). These techniques can be fundamentally classified into timing, concentration, and type; besides, some hybrid types of modulation will be described at the end of this Sect. 3.3.5.5.

3.3.5.1

Time Modulation

In the molecular communications via diffusion MCvD scenario shown in Fig. 3.22, a sequence of input symbols (or bits) are modulated in a time-slotted manner into M symbols [89]: .

S = [S1 , S2 , . . . , SM ]

(3.5)

3.3 Molecular Communication Systems

109

Fig. 3.22 Molecular communication system and its three core components: the transmitter (T x), the molecular communication channel, and the receiver (Rx)

where .SK ∈ S refers to the k-th symbol modulated onto a chemical signal by encoding the symbol into some properties of the chemical emission process and S is the symbol set. The modulated signal is transmitted through the MC channel, where the propagation environment degrades the signal and introduces a delay probabilistically. The channel-impaired chemical signal is detected and demodulated at the receiver. In Eq. 3.6, it is expressed as a set of received symbols such that .Sˆk refers to the k-th symbol demodulated and detected at the receiver. The detected symbols are fed to the demodulator that can correct some of the errors in detection to recover the information [89]: .

Sˆ = [Sˆ1 , Sˆ2 , . . . , SˆM ]

(3.6)

Over time, some information particles called messenger molecules (MMs) reach the Rx, which uses a detection process that depends on the type of the MM and the type of receiver used for detection. To date, the detection processes that dominate the MC literature are passive receivers and absorbing receivers. In passive receivers, the Rx acts like a transparent entity (i.e., as if the Rx is not in the environment) and does not affect the movement of the MMs. The detection consists of counting the concentration of the MMs inside the Rx at certain time intervals. In absorbing receivers, the Rx acts as an absorbing entity for the MMs, and when the MMs hit the Rx, they are removed from the environment. In these receivers, the detection can be modeled as counting the number of the MMs that hit the Rx within a given time interval. A subcategory of the absorbing receivers is the partially absorbing receivers, where only some parts of the receiver can absorb MMs while other parts reflect them [89].

110

3.3.5.2

3 Analysis of the Molecular Physical Layer’s Tasks

Concentration-Based Techniques

The main idea of concentration-based techniques is carrying information on the released MM concentration over discrete period time slots (i.e., symbol slots), where each slot is used to carry a single symbol of the overall message [90, 91]. In most works, each time slot has a fixed period. In its simplest form where each symbol represents a one-bit value (called on-off keying (OOK)), if the corresponding bit value .(S[k]) is bit-1, the T x releases a fixed number of MMs (i.e., .n1 ). On the other hand, if it is bit-0, the T x does not release any molecules for that symbol slot. At the receiver side, the Rx counts the number of MMs that arrive within each symbol slot (i.e., .N Rx [k]) and makes a threshold-based decision to decode the bit value of ˆ ˆ the given symbol slot .(S[k]). If .N Rx [k] ≥ λ, S[k] is decoded as bit-1; else, it is decoded as bit-0, where .λ is a threshold value for signal detection [89]. A more generalized version of OOK is called concentration shift keying (CSK) where, depending on the system design, each symbol represents m-bits of information. Following the classical modulation terminology, if a symbol represents one bit of information, this technique is called binary CSK (BCSK). If a symbol represents two bits of information, it is called quadrature CSK (QCSK) and so on. In CSK, for the k-th symbol in the message, the T x releases .N T x [k] number of MMs depending on the current symbol value as [89]: .

N T x [K] = nS[K] , S[K] ∈ {sym0 , sym1 , . . . , sym2m −1 }

(3.7)

where .nS[k] denotes the number of molecules to be emitted for the symbol value of .S[k] that can take one of the .2m symbol values (e.g., .sym0 , sym1 ) in the ˆ from the received signal, the Rx modulation alphabet. In order to demodulate .S[k] m uses .2 − 1 thresholds (i.e., .λ0 , λ1 , . . . , λ2m −2 as [89]): ⎧ N Rx [K] < λ0 ⎨ sym0 , ˆ .S[k] = symi , λi−1 ≤ N Rx [K] < λi , 1 ≤ i ≤ 2m − 2 ⎩ sym2m −1 , λ2m −2 ≤ N Rx [K]

(3.8)

where .N Rx [k] denotes the number of received molecules during the k-th symbol slot. In the literature, numerous CSK variants have been proposed to increase communication performance, e.g., to decrease the error probability of the system (Table 3.1). Most of the CSK variants consider the T x releasing all the MMs based on the selected .N T x [k] value at the start of the corresponding symbol slot [89, 92].

3.3.5.3

Type-Based Techniques

The second group of modulation techniques, called type-based techniques, focuses on using multiple types of MMs in the communication system as the basis of

Technique identification modulation characteristics References ISI mitigation Name CSK [93, 94] None [95] None CSK-AD CSK-SD None [96, 97] None [98, 99] CSK-sub TS High CSK-PA [100–102] CSK-CPA High [103] [104] Low CSK with ATD Moderate CSK with ML, MAP, MMSE [105, 106] NC-CSK-diff None [106] Moderate [107] NC-CSK-gamma Complexity Low Low Low Low High Moderate Low High Low Low

Performance evaluation assumptions Detection type Tx waveform Thresholding Pulse Inst. thresholding Pulse Thresholding Pulse Thresholding Imperfect pulse Thresholding Pulse Thresholding Pulse Adaptive thresholding Pulse ML, MAP, MMSE Pulse NC diff Imperfect pulse NC ML Pulse

Table 3.1 Concentration-based modulation techniques for molecular communication Rx type Absorbing Passive Passive Absorbing Absorbing Absorbing Absorbing Absorbing Passive Passive

Environment 3D, No drift 3D, no drift 3D, no drift 1D, with drift 1D–2D, no drift 1D–2D, no drift 3D, no drift 3D, no drift 3D, no drift 3D, with drift

3.3 Molecular Communication Systems 111

112

3 Analysis of the Molecular Physical Layer’s Tasks

the modulation technique. In these techniques, the transmitter has the capability of releasing different types of MMs .(mmtype ), which are similar to each other in composition (i.e., radius, diffusivity) but can only be received by a particular type of receptor at the receiver surface. Using different types of receptors, each corresponding to a particular type of MM, the receiver can receive multiple molecular release signals, which are practically orthogonal to each other, within a single symbol slot [90]. The quantitative feature of these orthogonal molecular release signals used to represent bit values depends on the particular technique in question [89, 108]. In the first type-based modulation technique, molecular shift keying (MoSK), each different symbol value .(S[k]) is represented by a specific type of MM. MoSK uses two types of MMs to modulate one bit of information within a symbol (called binary MoSK (BMoSK)) or four types of MMs to modulate two bits of information within a symbol (called quadruple MoSK (QMoSK)). Considering BMoSK, for each symbol slot, the Rx counts the number of arriving MMs for each MM type and demodulates S [k]. Based on the thresholding decisions for each MM type as [89]: ⎧ Rx Rx ⎨ sym0 , Nmma [K] ≥ λ ∧ Nmmb < λ Rx Rx ˆ .S[k] = sym1 , Nmmb [K] ≥ λ ∧ Nmma < λ ⎩ e, otherwise

(3.9)

where a and b represent the types of the MMs used, .MM1 ∈ (mma , mmb ), and R Nmm x[k] represents the number of .mmtype received at the Rx during the .k th type symbol slot [89]. An alternative approach in type-based techniques is to use a majority-based ˆ detection instead of thresholding. In such a binary type-based technique, .S[k] is decoded as [89]:

.

 ˆ S[k] =

.

Rx [K] > N Rx [K] sym0 , Nmm mmb a Rx [K] ≥ N Rx [K] sym1 , Nmm mma b

(3.10)

Besides the basic MoSK, other type-based techniques have also been proposed in the literature, each focusing on different methods of utilizing multiple MM types in representing the received molecular signal (Table 3.2) [89, 90]:

3.3.5.4

Timing-Based Techniques

When MM is released, information is encoded using this modulation technique. The channel input is basically continuous in this modulation as opposed to discrete in the preceding techniques [90]. In its simplest form, an MTC (molecular timing channels (MTC) is a subclass of MC channels) is based on a single MM released by the T x at

Technique identification modulation characteristics References Name ISI mitigation [94, 109–111] Moderate MoSK [112, 113] Moderate IMoSK MoSK with ML High [114] [115] RSK Moderate High MCSK [116] [101] MTSK High High [117] Pre-eq. CSK High [118] Zebra-CSK Complexity Moderate Moderate High Moderate Moderate High High High

Table 3.2 Type-based techniques for molecular communication Performance evaluation assumptions Detection type T x waveform Thresholding Pulse Thresholding Pulse ML Pulse Thresholding Pulse Thresholding Pulse Thresholding Pulse Thresholding (Diff) Pulse Thresholding Pulse

Rx type Absorbing Absorbing Passive Absorbing N/A Absorbing Absorbing Absorbing

Environment 3D, no drift 1D, drift 2D, no drift 1D, drift N/A 2D, no drift 3D, no drift 3D, no drift

3.3 Molecular Communication Systems 113

114

3 Analysis of the Molecular Physical Layer’s Tasks

timetr with information encoded on this release time .(RT ). The MM goes through some random propagation and arrives at the destination at time [89]:

.

.

ty = t r + tn

(3.11)

where .tn is some random delay due to MM propagation. Note that unlike previous modulations where the symbol set is finite, in pure timing channels, the symbol set is a continuous interval [89]. In Table 3.3, a variation regarding this modulation techniques is observed [89]:

3.3.5.5

A Hybrid Type of Modulation

• Example I In [135], the authors consider an MCvD link between a single-point transmitter (T x) and a single synchronized, spherical, and absorbing receiver (Rx) in an unbounded, driftless 3D environment. Thus, they propose a hybrid MCvD modulation scheme that utilizes a single molecule type. The proposed scheme combines PPM constellations with conventional CSK symbols and is called molecular concentration-position modulation (MCPM). The authors in [135] find that the selection of concentration constellations in MCPM poses a tradeoff between errors in detecting concentration or position-based constellations, suggesting an optimization problem on the emission intensity difference of the concentration constellations while keeping the average number of emitted molecules constant. The proposed scheme merges the conventional binary CSK (BCSK) with PPM constellations to construct a single-type molecule hybrid modulation scheme. At the transmitter side, the bit sequence u is arranged in groups of 1 + log2 k bits, where k is the PPM order that is to be employed. The first log2 k bits determine the position-based constellation, which is the release subinterval of the emitted molecules. After selecting the molecule emission’s temporal position, the group’s last bit decides on the BCSK symbol to modulate the intensity of the transmitted signal, resulting in a concentration-position joint symbol. Stemming from its components, the scheme is called k − ary molecular concentrationposition modulation (k − MCP M), where the parameter k defines the utilized k − ary PPM [135]. To provide a fair comparison among the error performances of MCvD schemes, different schemes are analyzed under the same bit rate (1/tb ) and the average number of emitted molecules per bit (M). While adopting the average emitted molecules per bit (M) and the communication bit rate (1/tb ) normalizations, k − MCP M can emit its symbols with a symbol duration of tsym = (1 + log2 k)tb and with (1 + log2 k)M molecules on average. Note that tsym denotes the total duration of a single-joint MCPM symbol. The bin (subinterval) durations of the utilized position constellations can be denoted as ts , which leads to tsym = kts [135].

Technique identification modulation characteristics References Name ISI mitigation [99, 119] None RT-single [120–122] None RT-multi RT-multi with Viterbi [123] None [124, 125] PPM Low PPM with ML [118, 126, 127] None [128, 129] PPM with decay None Low [130–133] Time between pulses None [84, 134] MFSK Complexity Medium High High Low Medium Medium Medium High

Table 3.3 Timing-based techniques for molecular communication Performance evaluation assumptions Detection type T x waveform ML Single N/A Multi ML/Viterbi Multi Thresholding Multi ML/max Multi First arrival/averaging Multi Thresholding/matched filter Multi/single Bandpass/matched filter Multi

Rx type Absorbing Absorbing Absorbing Absorbing Absorbing Absorbing Absorbing Absorbing

Environment 1D, drift 1D, no drift 1D, no drift 3D, no drift 1D–3D, drift 1D, no drift 1D–3D, drift 1D–3D, no drift

3.3 Molecular Communication Systems 115

116

3 Analysis of the Molecular Physical Layer’s Tasks

Noting the aforementioned constraints and assuming equiprobable bit-0 and bit-1 transmissions, the concentration difference between the bit-0 and bit-1 of the BCSK constellation in an MCPM scheme is a design problem. When emitting the molecules with 2 α (1 + log2 k)M and 2 (1 − α)(1 + log2 k)M for bit-1 and bit-0, respectively, the value of α needs to be optimized within the region (α ∈ [0.5, 1]) to ensure the system emits (1 + log2 k)M molecules on average. Note that if one were to select an α close to 0.5, the concentration constellations become very close to one another, which makes them harder to detect at the Rx end. On the other hand, an α value close to 1 would make it difficult for the Rx to detect the PPM symbols since if the bit that modulates BCSK is zero, close to zero molecules are emitted to the channel. To demonstrate the proposed scheme, the transmission strategy of an exemplary MCPM scheme is presented in Fig. 3.23. In addition, Table 3.4 presents the average emitted number of molecules per symbol and the different orders of MCPM, alongside conventional PPM and BCSK [135]. To detect the concentration-position joint symbols of the MCPM scheme, the detectors on the PPM and BCSK components of the scheme are employed consecutively. The overall MCPM detector that is utilized in this paper detects the ith joint concentration-position symbol, denoted by si , in two steps [135]: 1. The first step is to detect the PPM constellation based on the largest molecule count among the PPM bin intervals using an arg max operation on the appropriate subintervals. On the i th concentration-position joint symbol, the maximum count detector can obtain the detected PPM constellation,Jˆl , by performing Jˆl =

Fig. 3.23 Transmission strategy of 4-MCPM. The first two bits determine the emission time of the molecular signal, while the third and last bit determines the emission intensity of the signal

3.3 Molecular Communication Systems

117

Table 3.4 Slot durations and average emitted molecules per symbol for single-type molecule mcvd schemes Modulation scheme Transmitted bits per unit symbol Subintervals per symbol Subinterval durationb (ts ) Bit duration Molecules per symbol (M T X )

Molecules per bit (on average)

BCSK 1

2-PPM 1

4-PPM 2

8-PPM 3

2-MCPM 2

4-MCPM 3

8-MCPM 4

1

2

4

8

2

4

8

tb

(1/2)tb

(1/2)tb

(3/8)tb

tb

(3/4)tb

(1/2)tb

tb 2M for bit-1, 0 for bit-0

tb M

tb 2M

tb 3M

M

M

M

M

tb BCSK bit-1 2M x 2α BCSK bit-0 2Mx2 (1 − α) M

BCSK bit-1 3M x 2α BCSK bit-0 3Mx2 (1 − α) M

BCSK bit-1 4M x 2α BCSK bit-0 4Mx2 (1 − α) M

max arg j ∈{(i−1)k+1,...,ik} Rj . Note that this step is motivated from the assumption that h1 is the largest channel coefficient [135]. 2. Considering the Jˆith subinterval yields the detected PPM component of Sˆi , the MCPM detector performs a fixed threshold detector on the arrival count at the Jˆith subinterval to detect the concentration symbol. Denoting the hypotheses corresponding to Sˆi ’s BCSK component being a bit-1 or a bit-0 as H1 and H0 , respectively, the MCPM detector obtains the concentration symbol by comparing H1 it to a threshold γ as RJˆ ≷ γ . Note that the presented detector is memoryless, as H0 it only considers the K subintervals of the joint MCPM symbol it detects [135].

• Example II In [136], the authors consider a molecular communication system having multiple transmitters at the source and receivers at the sink bio-nanomachine in a 3D medium and propose a hybrid modulation scheme combining MoSK and CSK. The random locations of these source bio-nanomachines are modeled as uniform PPP (Poisson point process) in the 3D space outside the tagged receiver volume. In the proposed hybrid MoSK-CSK modulation-based system, each transmitter-receiver link in a source-sink pair uses M − ary CSK modulation with different transmitter-receiver links using different types of molecules. Thus, the modulation scheme consists of MoSK and M − ary CSK to form hybrid modulation. The use of different types of molecules across links completely avoids interlink interference at any receiver of a sink bio-nanomachine since the

118

3 Analysis of the Molecular Physical Layer’s Tasks

molecules from undesired transmitters of the paired source do not interfere. Due to the absence of ILI (interlink interference), the analytical expression for channel response can be applied to the system instead of just the numerical results via simulation-based approaches. Also, the molecules of the same type from other interfering sources arriving at the receiver reduce, and hence MUI (multi-user interference) also reduces, compared to that of MIMO systems using the single type of molecules for all links [136]. Thus, in [136, 137], a molecular communication system with multiple sources and one sink, where only one source randomly communicates to a sink and others act as interfering sources in a 3D homogeneous pure-diffusive fluid medium, is considered. In [136], the schematic diagram of a single source-sink pair is shown in Fig. 3.24. Each source-sink pair consists of a source bio-nanomachine having N point transmitters attached to it and a sink bio-nanomachine with N fully absorbing spherical receivers. The nth receiver has a radius rn and a volume Vn . A typical transmitter-receiver link between the nth transmitter and the nth receiver uses M − ary CSK modulation and uses molecules of type-n different from the molecules used by other transmitter-receiver links. As exposed in Fig. 3.24, the source first converts the serial information bits to N parallel bit sequences of length log2 M at each time slot. Then, each

Fig. 3.24 Schematic diagram of a source-sink pair, where different colors represent transmitterreceiver links using different types of molecules

3.3 Molecular Communication Systems

119

point transmitter emits molecules to the propagation medium according to M − ary CSK modulation. Therefore, the bit rate (Rb ) of this system is Rb = Nlog2 (M) per symbol period. At the sink, N spherical receivers with receptors that bind to only one type of molecule count the absorbed molecules. The acquired bit sequence is finally transformed to serial form after demodulation. Assuming perfect synchronization between transmitter and receiver pairs, the spherical absorbing receiver is able to count every molecule that is absorbed. With the exception of the associated receiver structures’ absorption, the source and sink are considered to be completely transparent to the diffusion of the signal molecules. Additionally, as was already indicated, this approach takes into account the fact that each receiver can only pick up the specific type of information molecule that its linked transmitter emits. This supposition is supported by the fact that a bio-restricted nanomachine’s size and complexity limit the types of molecules it can recognize. Furthermore, the elimination of ILI, which is common in systems with many transmitters and receivers that utilize a single type of information molecules, is a benefit of using distinct types of information molecules for separate transmitter-receiver linkages. Due to the absence of ILI, the suggested system’s intended separation between spherical receivers has no effect on performance [136]. The locations of the centroid of all sources are modeled as HPPP (homogeneous Poisson point process) in a three-dimensional (3D) space of R3 with density λ. Let Φj = xj i , i ∈ N be the point process representing locations of j th point transmitter TSi Xj i of all source Si . The ith sink is denoted by Ki and is paired with Si . Independent displacement of points in PPP according to some probability distribution will also be PPP according to the displacement theorem. Therefore, Φj is a PPP. The authors in [136] contemplate a typical pair S0 − K0 where a jth receiver RK0 ,0 of sink 0 is considered at the origin. For this tagged receiver, the Φj can be modeled as PPP outside the receiver (R3 \Vj ) with intensity λ. The desired transmitter TS0 ,xj 0 ) of the desired source S0 is a part of the PPP Φj , as allowed by Slivnyak theorem without affecting the results [136]. The research previously mentioned is mandatory to analyze the degradation of a molecular signal as follows: For a point source located at a distance r away from the center of a spherical receiver of radius rn , the hitting rate of molecules at the receiver is given by the next equation: .

q(t  |r) =

rn √ r−rn r 4π Dt 3

  2 n) exp − (r−r 4Dt 

(3.12)

where D denotes the diffusion coefficient, which depends on the molecule’s characteristics and the fluid environment. The detection performance can be improved by introducing an adequate amount of molecular degradation. The molecule should reach the receiver boundary before its degradation, and it should not reside in the environment for more than symbol time to avoid ISI. The rate of degradation of molecule (μd ) depends on the half-life (∧1/2 ) of molecule undergoing degradation, i.e., μd = ln(2)/∧1/2 [136].

120

3 Analysis of the Molecular Physical Layer’s Tasks

The fraction of nondegraded molecules absorbed by the spherical receiver within an arbitrary time t since transmission time can be obtained as [136]: f (μd , t|r) = =

.

1 rn 2 r

t 0

q(t´|r)e−μdt´ dt´



exp − μDd (r − rn )

 

   √ √ ∗ exp 2 μDd (r − rn ) erf c √r−rn + μd t + erf c √r−rn − μd t 4Dt

4Dt

(3.13) The above equation shows that the fraction of molecules absorbed at the spherical receiver reduces with an increase in μd . For the scenario with no molecular degradation, i.e., μd = 0, the above expression simplifies as [136]: .

f (0, t|r) =

rn r

 erf c

r−rn √ 4Dt

(3.14)

For the analysis of the channel model, as stated earlier, it is essential to mention that each of the transmitter-receiver links uses M −ary CSK modulation for communication. At time instant k, the j th (1 ≤ j ≤ N ) transmitter TSi ,xj i of j

the ith source machine Si randomly located at xj i = x emits ux [k] = Qm , 0 ≤ m ≤ M − 1 molecules at the beginning of symbol period Ts corresponding to j the message symbol SX [k] = Sm . For example, in OOK modulation, the j th transmitter emits either no molecules (Q0 = 0) or Q1 number of molecules corresponding to bit-0 or bit-1 [136]. The authors in [136] consider a typical j th receiver of pair S0 − K0 , located at the origin. They assume a discrete-time channel model with channel memory j of length L, where hx [l] denotes the channel impulse response at lth time instant for the channel between the typical receiver and the jth transmitter of the ith source Si which is randomly located at x according to uniform PPP. The CIR at the lth time instant, which is the fraction of molecules absorbed between lTs j and (l + 1)Ts , can be obtained from Eq. 3.13 and is given as hx [l] = f (μd , (l + 1)Ts | ||x||) − f (μd , lTs ||x||) [136]. Let the jth transmitter of the desired pair source S0 be located at x ∗ . Considering the hitting of information molecules on the receiver as success events, the number of molecules that were transmitted from TS0 ,X and detected by the j th spherical receiver at kth time instant follows binomial distribution j j with parameters (ux [k − l], hx [l]), which is approximated to Poisson distribution j assuming the number of information molecules ux [k − l] is large and the hitting j probability hx [l] is small. It is indispensable to note that the sum of independent Poisson random variables follows the Poisson distribution. Thus, the total number

3.3 Molecular Communication Systems

121 j

of molecules received at the j th receiver in kth time instant y0 [k] follows L hj [l]uj [k − l] [136]. That is: Poisson distribution with parameter x∈Φj l=0 x x ⎛ ⎞ I SI MU I Desired       ⎟ ⎜ j j j j j j j L L ⎟ .y [k] ∼ P ⎜h ∗ [0]u ∗ [k] +  l=1 hx ∗ [l]ux ∗ [k − l] + x∈Φj \{x ∗ } l=0 hx [l]ux [k − l]⎠ x 0 ⎝ x (3.15)

where P (.) represents Poisson distribution. The number of molecules received at the tagged receiver located at the origin is the sum of molecules corresponding to the desired symbol, previous symbol molecules which cause ISI, and molecules of the same type from other interfering transmitters, which result in MUI [136]. For the analysis of the probability of symbol error analysis, for simplicity, in [136], it is considered that all the j th transmitters of all sources of multiple pair system are sending the same symbols s j [k] with a probability of sending Sm as PSm at any time instant k. The sink bio-nanomachine uses fixed threshold detection in which, at every symbol period, the counted number of molecules is compared with the predetermined threshold values (T0 , T1 , . . . , TM ) and is decoded as Sm if the received molecule count is between Tm and Tm+1 . Let the j decoded symbol at the kth time instant be sˆ j [k] and Pm,se [k] = P j (ˆs j [k] = Sm | s j [k] = Sm , s j [k − L − 1 : K − 1]) at the kth time instant be the probability of an event that the sink fails to decode sˆ j [k] = Sm corresponding to the tagged receiver j given that the transmitted symbol is s j [k] = Sm and the previous j symbols s j [k − L − 1 : K − 1]. The probability of symbol error Pse [k] during time instant k at tagged receiver conditioned on L previous symbols is given by Sabu et al. [136]: j

.

M−1 Pse [k] = m=0 Psm P j (ˆs j [k] = Sm | s j [k] = Sm , s j [k − L − 1 : K − 1])    j

Pm,se [k]

(3.16) The likelihood of a symbol error versus different kinds of molecular modulation is shown in Figs. 3.25 and 3.26 using the methodology outlined above and taking into account a system with numerous sources and one sink. • Example III A MIMO-MC (multiple-input multiple-output molecular communications) system in a 3D unbounded environment with point transmitter sources and spherical receivers, which are presumed to be memoryless, is considered in [138]. In this MIMO-MC, perfect synchronization is also expected. The authors of [138] take into account a diffusion-based MIMO-MC system NxN that is operated at the microscale, where the transmit and receive N and N nanomachines are, respectively, linked to the transmitter and receiver sides of the cell membrane. Additionally, they presuppose that the centers of the transmitter and

122

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.25 Symbol error probability versus distance between point transmitter and spherical receiver (r − rj ) in a system with single link [136]

Fig. 3.26 Probability of symbol error versus threshold T1 for the detection of symbols 0 and 1

receiver cells, respectively, serve as transmission control centers and reception decision centers. While the reception decision center connecting all the receiver nanomachines decodes the information based on the received signals from receiver nanomachines, the transmission control center’s job is to coordinate the transmitter nanomachines to emit molecular pulses in accordance with the information to be transmitted. Figure 3.27, which depicts the lipid bilayer of a cell membrane in green, is an example of a MIMO-MC system with pairs of 4 × 4 transceivers used to illustrate the communication system model of the MIMO-MC. The transmitter and receiver of a link, as well as the molecules sent between them, are indicated using the same color for simple identification, as seen in Fig. 3.27. Additionally, in [138], it is necessary to consider the following

3.3 Molecular Communication Systems

123

Fig. 3.27 System model of MIMO-MC

notation: Boldface uppercase and lowercase letters indicate matrices and vectors, respectively. Rn×m indicates a real valued matrix with n × m dimensions. E[·], || · ||, | · |, Q(·) and Pr [·] represent expectation, Euclidean norm, absolute value, Q-function, and probability of an event, respectively. As is common knowledge, ILI and ISI occur in MIMO-MC, which may considerably reduce the performance of the detection system. The SM-MC (spatial modulation for molecular communications) is suggested in [138] as one of the implementations of the MIMO-MC to address these issues. This method uses a single type of molecule to transmit information while concurrently utilizing the spatial and concentration domains. The SSK (shift space keying) techniques are the source of inspiration for the SM-MC philosophy. Generally, during each symbol period in SSK modulation, only one transmit antenna is turned on. When the channel status information is provided, the receiver can determine the index of the transmit antenna that is activated. A typical amplitudephase modulation can be used to implement the SSK modulation, creating the SM. A CSK modulation and an SSK modulation are combined in the SM-MC as suggested in [138]. One of the transmitter nanomachines emits a pulse of molecules in response to a communicated symbol, with the number of molecules released also depending on the data symbol being transmitted. In Fig. 3.28 [138], the transmit schematic diagram for the aforementioned SM-MC system is shown in detail.

124

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.28 Transmitter diagram of SM-MC

Hence, when the j -th space symbol is transmitted, with j ∈ {1, 2, . . . , N }, the transmit signal vector has a form of [138]: ⎡ .

⎤T

⎢ x(t) = ⎣0, 0, . . . ,

S m 

⎥ , . . . , 0, 0⎦

(3.17)

j -th transmitter

where only the j -th transmitter is activated to emit Sm > 0 molecules, when the m-th concentration symbol Sm is selected. The space symbol is denoted as S j when the j -th transmitter is activated. In [138], it is assumed that the space symbol S j and the concentration symbol Sm are independent of each other, solely depending on the input data stream; then [138]: .

P r[Sm ] =

1 M,

m ∈ {0, 1, . . . , M − 1}

(3.18)

j ∈ {1, 2, . . . , N}

(3.19)

and .

P r[S J ] =

1 N,

In [138], it is also assumed that the concentration vector corrupted by noise sensed by the receivers at time t can be expressed as [138]:

3.3 Molecular Communication Systems

.

125

yi,MI MO (t) = hii (t)Xi (t) +    desired signal

+ ni,MI MO (t)   

Ii,MI MO (t)    sum of interf erence

(3.20)

noise

Regarding the transmitter, SM-MC in [138], if the SM symbols represent the j combinations of the space and concentration symbols as Sm , then [138]: j

P r[Sm ] =

.

1 j ∈ {1,2,...,N } , MN m ∈ {0,1,...,M−1}

(3.21)

Therefore, the raw data rate of the SM-MC measured in bits per molecular symbol is given as [138]: .

RSM = log2 N + log2 M

(3.22)

where both the values of N and M are assumed to be an integer power of 2. Based on Eq. 3.17, the concentration signal observed at the i-th receiver in the SM-MC systems is similar to Eq. 3.20 and can be expressed as [138]: .

yi,SM (t) = Sm hij (t) + Ii,SM (t) + ni,SM (t)

(3.23)

where both Sm hij and ni,SM (t) are dependent on the current molecular symbol being received, representing the expected number of molecules received and the noise component, respectively, at time t, when the j -th transmitter is activated to emit a chemical impulse with Sm molecules. However, unlike the case in MIMOMC, the Ii,SM (t) in Eq. 3.23 consists of only the ISI component that resulted from a previous molecular symbol emitted by the i-th transmitter, since in [138] it is assumed that ILI only occurs with the current transmission. At the same time, ISI is only experienced from one previous emission by a paired transmitter. J¯ , then the interference component Denoting the previous molecular symbol as Sm ¯ is [138]: * Ii,SM (t) =

.

0, f or i = j, Sm¯ hii (t + Ts ), f or i = J¯

(3.24)

The distribution of the noise component in Eq. 3.23 can be expressed as [138]: .

2 ni,SM (t) ∼ N(μni, SM (t), σni, SM (t))

2 associated with μni, SM (t) = 0, σni, SM (t) =

Sm hij (t)+Ii, VRX

(3.25) SM (t)

[138].

126

3 Analysis of the Molecular Physical Layer’s Tasks

Furthermore, based on Eqs. 3.23 and 3.25, the distribution of yi,SM (t) can be found and expressed as [138]: .

 2 yi,SM (t) ∼ N Sm hij (t) + Ini, SM (t), σni, SM (t)

(3.26)

The concentration vector sensed at the N receivers in the SM-MC system is expressed as: .

+ ,T ySM (t) = y1, SM (t), . . . , yi, SM (t), . . . , yN, SM (t)

(3.27)

From the above analysis and Eq. 3.24, authors in [138] can readily realize that their proposed SM-MC transmission scheme is capable of getting rid of the most significant ILI generated by the signals sent by the unpaired transmitters during the current symbol period, which is unavoidable in the general MIMO-MC. Furthermore, the ILI generated by the previous symbols can also be significantly mitigated owing to the employment of SSK modulation. The molecular energy efficiency is defined as [138]: .

η=

R(1−Pe ) S¯

≈

R¯ S¯

(3.28)

whose unit is bits per molecule. This definition corresponds to the concept of energy efficiency in wireless communications, measured in bits per J oule. Pe is the BER, S¯ is the mean number of molecules released for each transmission, R is the raw data rate, and R¯ is the achievable rate. Simulation results demonstrate that the SSK-MC and SM-MC schemes can attain lower Pe than the conventional ¯ Hence, SSK-MC and SM-MC can obtain MIMO-MC under the same R and S. higher molecular energy efficiency than the conventional MIMO-MC. The raw data rates R of some modulation techniques for MC are illustrated in Table 3.5 [138]. The reason behind the above results can be interpreted in detail as follows: For an N ×N MIMO-MC with a fixed modulation scheme, increasing N may achieve a higher raw data rate but at the expense of the degraded error performance. ¯ This is because the number of molecules released per transmitter is S/N, which reduces as N increases. Second, a larger N infers that ILI is imposed by more undesired links, which further degrades the error performance. By contrast, for Table 3.5 Ideal data rate for different modulation schemes in MC

Scheme OOK BMoSK 2 × 2 MIMO (OOK) 2 × 2 SM (BCSK) 4 × 4 SM (BCSK) 2 × 2 SM (QCSK)

Ideal data rate 1 1 2 2 3 3

Scheme BCSK BSSK QCSK QSSK 8SSK 16SSK

Ideal data rate 1 1 2 2 3 4

3.3 Molecular Communication Systems

127

the SM-MC and its special scheme of SSK-MC, owing to their capability of ILI mitigation, they can obtain a logarithmically increasing raw data rate. This is solely contributed by the space domain degrees of freedom without demanding extra transmission energy (molecules) and introducing additional ILI. Therefore, authors in [138] may argue that the SSK-MC and SMMC schemes are suitable for the microscale communication scenarios where the energy supply is limited. Regarding SM-MC signal detection, theoretically, the ML (maximum likelihood) that is an optimal detection performance can be implemented with the SM-MC to decode the space and concentration symbols jointly. When memoryless receivers are considered, the optimal ML detection based on Eqs. 3.23 and 3.27 can be formulated as [138]: .

 jˆ, m ˆ =

argmin 2 j ∈ {1, 2, . . . , N } , m ∈ {0, 1, . . . , M − 1}||ySM (t)−Sm hj (t)|| (3.29)

where Jˆ and m ˆ are the estimated indices of the space and concentration symbols, respectively. Note that Sm in SM-MC is always greater than zero, i.e., Sm > 0; otherwise, the space symbols in the case of Sm = 0 are unable to be detected. Equation 3.29 infers that the ML detector has a search complexity of O(N M). Explicitly, when NM is relatively large, it is not practical to be deployed with MC systems, owing to the constraint on the size and computing capability of MC receivers. Therefore, low-complexity detection schemes are desirable for MCs. To this end, [138] proposes a low-complexity successive detection scheme for the SM-MC. In this scheme, the CSK symbol is detected after detecting the space symbol as follows: First, to be more specific, the index of the activated transmitter is detected via comparison of the concentration sensed by the N receivers, based on the fact that the receiver paired with the activated transmitter is most likely to have the maximum concentration at the sampling time because it is located with the minimum distance from its paired active transmitter. This is a noncoherent detection scheme, which can be formulated as [138]: .

argmax Jˆ =j ∈{1,2,...,N } yj, SM (t)

(3.30)

After the detection of the space symbol, the concentration observed by all the N receivers can be exploited for detecting the CSK symbol, which can be formulated as [138]: argmin

m ˆ =m∈{0,1,...,M−1} ||w T ySM (t) − Sm w T yj (t)||2

.

(3.31)

where w ∈ RN x1 is a weighting vector of length-N, expressed as [138]: .

w = [ω1 , . . . , ωi , . . . , ωN ]T

(3.32)

128

3 Analysis of the Molecular Physical Layer’s Tasks

which may be different when different combining strategies are employed. In [138], it proposed and investigated two types of combining strategies that are commonly used in wireless communications, which are selection combining (SC) and maximal ratio combining (MRC). Specifically, when the SC-assisted detector is employed, only the Jˆ-th receiver’s concentration, where Jˆ is obtained from the space symbol detection, as seen in Eq. 3.30, is utilized to detect the concentration symbol. In this case, the components in Eq. 3.32 are given the values as [138]: * ωi =

.

1, f or i = Jˆ, 0, otherwise

(3.33)

Then, upon applying these results to Eq. 3.31, the CSK symbol index detected using the SC-assisted detection can be described as [138]: argmin

m ˆ SC =m∈{0,1,...,M−1} |yJˆ, SM (t) − Sm hJˆJˆ (t)|2

.

 2 argmin = m∈{0,1,...,M−1} yJ2ˆ,SM (t) − 2Sm yJˆ, SM (t)hJˆJˆ (t) + Sm hJˆJˆ (t)  2 (a) argmin = m∈{0,1,...,M−1} Sm hJˆJˆ (t)

− 2Sm yJˆ, SM (t)hJˆJˆ (t)

(b) argmin 2 = m∈{0,1,...,M−1} (Sm ) hJˆJˆ (t) − 2Sm yJˆ, SM (t)

(3.34)

where (a) to (b) is due to hJˆJˆ (t) being a real positive number. From Eq. 3.34, authors in [138] can readily know that the complexity of our proposed SCassisted detector is O(N +M). Provided that N > 2 or M > 2, the complexity of the SC-assisted detector is lower than that of the optimal ML detector of Eq. 3.39. When the MRC detector is employed, w = hJˆ (t). Correspondingly, the CSK symbol index can be estimated from the optimization problem of [138]: -2 argmin m ˆ MRC =m∈{0,1,...,M−1} -hTJˆ (t)ySM (t) − Sm ||hJˆ (t)||2 argmin

=m∈{0,1,...,M−1} ||hTˆ (t)ySM (t)||2 − 2Sm hTˆ (t)ySM (t)||hJˆ (t)||2 J

.

+ (Sm ) ||hJˆ (t)|| 2

J

4

  argmin =m∈{0,1,...,M−1} ||hJˆ (t)||2 (Sm )2 ||hJˆ (t)||2 − 2Sm hTJˆ (t)ySM (t) argmin

=m∈{0,1,...,M−1} (Sm )2 ||hJˆ (t)||2 − 2Sm hTˆ (t)ySM (t) J

(3.35)

3.3 Molecular Communication Systems

129

Alternatively, after the space symbol is detected, if it used the ML detection to detect the CSK symbol, the detection problem can be described as [138]: argmin

m ˆ Sub−ML =m∈{0,1,...,M−1} ||ySM (t) − Sm hJˆ (t)||2 argmin

T =m∈{0,1,...,M−1} ||ySM (t)||2 − 2Sm ySM (t)hJˆ (t) .

+ (Sm )2 ||hJˆ (t)||2 argmin

T =m∈{0,1,...,M−1} (Sm )2 ||hJˆ (t)||2 − 2Sm ySM (t)hJˆ (t)

(3.36) T (t)h (t) = hT (t)y Intuitively,ySM SM (t). Hence, the detection problem of Jˆ Jˆ Eq. 3.35 is the same as that of Eq. 3.36, meaning that the MRC-assisted detector is equivalent to the ML detector for detecting the CSK symbol. Hence, MRCassisted detection is also optimum. From Eq. 3.32, we can obtain the complexity of the MRC-assisted detector is also O(N + M), although the total number of computations of Eq. (3.35) is significantly higher than that of Eq. 3.34 [138]. • Example IV For a diffusion-based MC system, the authors of [139] offer a novel binary bit addition modulation technique that makes use of the straightforward binary addition concept and lowers the actual amount of molecules that are released from the transmitter. Here, it is referred to as a binary bit addition modulationbased molecular communication through diffusion (MCvD) communication system. The geometry of a physical transmitter is unaffected since the transmitter is a point or is portrayed as a zero-dimensional point. Moreover, it is presumable that molecules are created instantly and enter the physical channel immediately. It is believed that the receiver is a completely absorbing receiver, meaning that the various receptors, each of which is sensitive to a different type of molecule, immediately absorb the signaling molecules of type 1 or 2 that arrive at the receiver by diffusion. Moreover, it is believed that the molecular noise is signaldependent, having a Gaussian distribution for the mean and variance that are connected to the signal [139]. In [139], the proposed modulation scheme is implemented using a simple binary adder; for instance, if the symbol ‘11 which is to be transmitted is input to the binary adder, it returns a carry of 1 and the sum as 0. As a first block, the transmitter consists of the molecule allocator (MA). It consists of two binary adders, with the sum output of the first binary adder being the input of the second binary adder. The output of the MA is the carry output of the two binary adders and the sum output of the second binary adder. The sum output of the first binary adder is expressed as S1 = b1 ⊕ b2 where b1 and b2 denote the binary inputs to the first adder and ⊕ denotes the binary addition and the carry output is given as C1 = (b1 .b2 ). Similarly, the output of the second adder is expressed as S2 = b1 ⊕ S1 [139].

130 Table 3.6 Binary bit addition modulation technique depicting molecule transmission scheme

3 Analysis of the Molecular Physical Layer’s Tasks Symbols 00 01 10 11

Molecule type T1 T1 T1 T2

Molecule count 0 L1 L2 L1

C1 0 0 0 1

C2 0 0 1 0

S2 0 1 0 0

The transmitter consists of MA as the first block whose outputs are C1 , C2 , and S2 , based on which the number of molecules to be transmitted is decided. The molecules of type 1 or type 2 are released based on which of these outputs of the MA is high. For instance, when C1 is high, L1 molecules of type 2 are released, and when S2 is high, L1 molecules of type 1 are released. To change the transmitted molecule from type 1 to 2, a carrying value of “1” is needed. Table 3.6 depicts the molecule transmission scheme for the proposed binary bit addition modulation technique. The total number of molecules for all the symbols is expressed as 2 x L1 + L2 . The binary bit addition modulation scheme can be considered a combination of BCSK and binary molecule shift keying (BMoSK) technique. If the input bit symbol is “00”, no molecule of type 1 or type 2 are released; if the input bit symbol is “01,” L1 molecules of type 1 are released; if the input bit symbol is “10,” L2 molecules of type 1 are released; and if the input bit symbol is “11,” L1 molecules of type 2 are released [139]. We use two receptors at the receiver to pick up the two different transmitted chemical kinds. In the receiver, a threshold is used to compare the molecules that were received. The receiver that is being used is totally absorbing. Type 1 and type 2 signaling molecules that diffuse to the receiver surface are immediately absorbed. The sensing area is made up of every point on the receiver’s surface, and the signal received is the total number of molecules that were absorbed in a brief period of time. Moreover, in the traditional BCSK modulation technique, just the molecule’s concentration is changed, resulting in a different amount of molecules for various signals. In the suggested modulation method, the symbol “11” is represented by a different kind of molecule than other symbols, which significantly reduces interference brought on by the same molecules from earlier slots [139]. The performance of the method planned in [139] defines the proposed modulation scheme with the two-bit BCSK modulation scheme and performs Monte Carlo simulations with 106 iterations. To demodulate the received signal, i.e., the number of molecules received, the threshold needs to be fixed; in this case, the authors vary the threshold from zero to the number of molecules transmitted, and the threshold that yields the lowest BER is chosen as the threshold for the specific parameters. This is termed as the adaptive threshold, which changes for different parameters. The distance between the emitter and the receiver and the radius of the receiver are taken as 2 μm for all the simulations. Authors assume the diffusion process to be one-dimensional. To get the simulation results, it is assumed that the transmitted bit sequence is equiprobable for all symbols. Large numbers of information molecules were obtained for the proposed modulation

3.3 Molecular Communication Systems

131

scheme analysis, and the Gaussian model was used to simulate the molecules that were really observed. Type 1 and type 2 molecules are released by the transmitter. The receiver’s threshold, which is in the range of hundreds of molecules, supports the assumption of the Gaussian model for the scenario under consideration, which is signal-dependent. When the number of transmitting molecules declines, the number of received molecules and the threshold also decline; the Gaussian assumption might not hold true [139]. In addition, the suggested modulation method accumulates less molecules in the following slots as the number of molecules rises than the two-bit BCSK scheme, where the interference rises as the number of molecules rises. According to the simulation results in terms of BER, the main benefit of the suggested modulation method is in situations where the number of molecules being communicated is high. Due to the enormous number of molecules received, they employ the Gaussian model for simulation [139]. The findings indicate that the proposed modulation scheme’s cumulative number of molecules received is more than those of the BCSK scheme and the typical signal realization. Comparing the initial and average number of molecules received at the receiver, we can see that, when compared to BCSK, the average value and initial realization of the proposed modulation scheme are closer in terms of the number of molecules [139]. • Example V It is recommended to apply the modulation techniques previously discussed to biological communications in [59]. The transmitter then indicates that information is encoded in this kind of communication by producing particles (or using stored particles). These particles typically range in size from a few nanometers to a few micrometers. These might be artificial substances like gold nanoparticles or biological substances like proteins or DNA molecules. The numerous ways in which information can be encoded (or modified) on the particles are depicted in Fig. 3.29. First, the intensity or concentration of particles can be used to encode (or regulate) information. For instance, the release of three particles can represent bit-0, while the release of one particle can represent bit-1. Second, different kinds of particles can be released in order to encode the information. For instance, type A particles are released to transmit bit-1, and type B particles are released to represent bit-0. It should be noted that a significant quantity of data can be conveyed in this manner using the structure of molecules. For instance, there are theoretically 432 distinct symbols that can be communicated in a 32 base pair single-stranded DNA sequence [59]. Ultimately, when particles are released, information can be encoded. By way of illustration, releasing the particles at the start of the symbol duration can stand for bit-0, while releasing them in the middle can stand for bit-1. Thus, the transmitter needs a way to regulate particle release [59]. Because it is challenging to regulate this process at the molecular level, the transmitter may malfunction. The information particles discharged go over a small space gap to the receiver. The aqueous or gaseous environment in which the transmitter and receiver are located allows for the unimpeded propagation of

132

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.29 Information can be encoded in the concentration or number of particles released, in their type or structure, or at the time of release

the tiny informational particles. The information is encoded when the particles are collected, detected, and detected at the receiver. Of course, there are a ton of specifics and variants on the concept. The “gap” (channel environment) could, for example, be a medium through which particles diffuse stochastically or active transport could be used. • Example VI A novel modulation technique for synthetic MC systems using photochromic signaling molecules is put out in [140]. This embeds the information into the photochromic molecule’s state. They take into account the green fluorescent protein variation “Dreiklang” (GFPD), whose fluorescence may be reversibly switched by light stimuli of mutually distinct wavelengths, as photochromic signaling molecules. Fluorescence, which refers to a molecule’s capacity to absorb light initially before reflecting it back at a higher wavelength and with reduced energy, enables the reading out of a GFPD molecule’s current state. The application of fluorescence has gained significant interest in designing synthetic MC systems over the last few years. In the experimental MC system in [140], a fluorescent dye was employed as an information carrier. To facilitate long-range information transfer in MC systems, fluorescent carbon quantum dots were investigated as signaling particles due to their concentration-dependent emission properties. In [140], the authors mention that a bacterial receiver (Rx) was proposed, which produces green fluorescent protein upon the reception of signaling molecules. Similarly, the authors in [140] mention the employment of genetically engineered Escherichia coli (E. coli) bacteria to exhibit fluorescence upon the reception of specific signaling molecules, thereby demonstrating that a chemical transmit signal can be converted to a fluorescence signal at the Rx. Also, in [140], an MC system for generating pulse-shaped signals with an NOR logic operation in engineered E. coli bacteria was proposed. Hereby, the yellow fluorescent protein’s fluorescence was employed as the NOR gate’s pulse

3.3 Molecular Communication Systems

133

signal. Besides, in [140], it mentioned researches which developed a multipleinput multiple-output (MIMO) nanocommunication system, where information transmission was based on the exchange of energy levels between fluorescent molecules employed as transceiver nano-antennas. According to [140], most existing works employ fluorescent molecules because of their easy detectability by the Rx. In the MC system proposed in [140], they propose fluorescent and nonfluorescent GFPD molecules correspond to state A and state B, respectively. The fluorescence of GFPD can be switched on (B → A) and off (A → B) by light stimuli at different wavelengths. Consequently, assuming the GFPD molecules are suspended in a fluid medium as elements of the envisioned synthetic MC system, optical sources emitting light at different wavelengths can be used as T x unit to modulate information (B → A) and eraser (EX) unit to delete the information (A → B). Moreover, a fluorescence detector equipped with a fluorescence-stimulating light source and an optical sensor can be employed as Rx [140]. • Example VII Ratio shift keying (RSK) modulation was initially thought of as encoding information into the concentration ratio of transmitted isomers that differ in the number of constituent monomers, as it was proposed in [141]. The same concentration ratio can then be achieved with various absolute concentrations of various molecule kinds, suggesting that there are additional options for efficient information flow across molecules. Second, assuming the impact of these variations is molecule-type invariant, RSK may be more resistant to dynamic changes in transmit power and channel impulse response (CIR). The mobile MC situation, which occurs when the diffusion coefficients of various molecular types are identical, serves as an illustration of this. The concentration ratio in the received signal would be preserved, and the time-varying CIR resulting from the mobility of the transceivers would be the same for both types of molecules at all times. In situations where the channel includes particular enzymes that break down both types of molecules at the same rate and do not alter the received concentration ratio, RSK can also be reasonably resilient. In a similar vein, scenarios where the transmitter can exhibit time-varying transmission patterns in terms of the absolute amount of molecules communicated while having a finite reservoir of molecules or variable molecule creation or harvesting methods are taken into consideration. The RSK can maintain its dependability if the transmitter can keep the transmitted concentration ratios under such circumstances. All of the aforementioned benefits of RSK, however, depend on the receiver’s ability to accurately detect the communicated concentration ratios. In that experiment, the effectiveness of RSK modulation was evaluated with respect to a physically relevant receiver design that is furnished with a single type of ligand receptor engaging cross-reactively with several types of molecules or ligands. The receiver can determine the received concentration ratio using maximum likelihood (ML) techniques by taking advantage of the differences in

134

3 Analysis of the Molecular Physical Layer’s Tasks

the affinities of the various types of ligands with the receptors, which are reflected in differences in receptor-ligand bound time duration statistics [141]. In order to assess the performance of the RSK modulation, we do a theoretical information analysis. The capacity of an end-to-end MC channel with a receiver using an optimal or suboptimal ratio estimate method is calculated analytically. The numerical results are compared to the capacity of the MC channel using a somewhat more traditional CSK modulation for different system settings, such as the similarity between ligand types used for ratio encoding and the number of receptors. According to the results, end-to-end MC channels with RSK exhibit a similar capacity to those with CSK, but they perform much better when the transmit power (or the highest concentration the transmitter can broadcast) is constrained. These findings point to RSK’s potential in the presence of time-varying channel and transceiver circumstances and suggest that it may be preferable to CSK and maybe other modulation methods for the development of molecule-efficient MC systems. The effectiveness of the suboptimal estimator, which is shown to be very similar to that of the ideal method, also suggests that the benefits of the RSK modulation can be attained with little computational complexity through biologically pertinent mechanisms [141]. • Example VIII In [142], it considers a uniformly separated array of transmitter (T x) and receiver (Rx) antennas embedded circularly as uniform circular array (U CA) on an infinite planer surface of T x and Rx units, respectively. The topology under consideration involves MoMIMO (molecular-MIMO) system with point sources as T x antennas capable of transmitting messenger molecules into the threedimensional, unbounded, diffusion-without-drift environment. The Rx surface consists of perfectly absorbing spherical Rx antennas aligned and synchronized with their respective point T x antennas. Each Rx antenna counts the number of molecules absorbed after the symbol period given by ts . In [142], the number of T x antennas (Nt ) and the number of receive antennas (Nr ) equal to 8, i.e., Nt = Nr = 8, are proposed. For simplicity, that work analyzed the STMCM (spatiotemporal molecular coded modulation) scheme for the 8 x 8 DBMoMIMO (diffusion-based MoMIMO) system. The distance between the nearest points of T x antenna and the corresponding spherical Rx antenna is denoted by dtr . Moreover, the distance between the closest point of Rx antenna from the center of the UCA is given by drc , and the radius of the Rx antenna is denoted by rr , respectively [142]. The random motion of information particles through free diffusion is realized using Monte Carlo simulations where the position of each molecule with diffusion coefficient D is updated after a time step of t = 10−4 . Since the DB-MoMIMO channel contains memory, the channel memory length L = 4 is considered in [142]. The channel characterization considers the dependency in the molecular arrivals at different spherical Rx antennas. In [142] is possible approximate the number of absorbed molecules as independent binomial random variable with the probability of success denoted by the i th channel coefficients

3.3 Molecular Communication Systems

135

pm , n[i] where 1 ≤ i ≤ L, and both m and n, varying from 0 to 7, denote the Tx and Rx antenna index, respectively. Since the MC system possesses rather primary capabilities, it has been shown in several works that MIMO approaches can significantly improve the performance of MC systems. MSSK (molecular space shift keying) scheme is shown to alleviate the problems of ISI and ILI in the DB-MoMIMO scenario. Despite robust schemes, ISI and ILI issues still prevail and impact the performance of DB MoMIMO systems. For the considered 8x8 DB-MoMIMO system, the first channel coefficient {i = 1}, i.e., pm,n [1] for the mth transmitter antenna and the nth Rx antenna, corresponds cth to the level ILI channel component with [142]: ⎧ ⎨ max(m, n) − min(m, n) ∀m, n = {0, 2, . . . , 7} , m = n, max(m, n) ≤ 4 .c = (min(m, n) − max(m, n)) mod8, ∀m, n = {0, 2, . . . , 7} , m = n, ⎩ max(m, n) > 4 (3.37) where 1 ≤ c ≤ 4. For instance, if the antenna constellation is numbered from {0, . . . , 7}, p0,3 [1] and p0,5 [1] correspond to the three-level ILI channel coefficients, and p0,1 [1] and p0,7 [1] correspond to the one-level ILI channel coefficients [142]. The primary factor in assessing system performance is the interference of currently transmitted messenger molecules with solely previously transmitted molecules. As a result, the suggested STMCM mapping approach in [142] is based solely on the antenna index that was previously activated. In light of the previously activated antenna index, the suggested approach selectively activates an antenna. Firstly, the transmitted bit stream is first divided into groups, each containing log2 (Nt /2) = 2 bits, where the total number of groups is equal to the total length of the transmitted symbol stream, i.e., DL . Each group having log2 (Nt /2) = 2 uncoded bits is judiciously coded into log2 (Nt = 3) bits, which are finally mapped to a particular T x antenna index to form the encoded symbol stream. Considering that the antenna constellation is indexed from {0, . . . , 7}, the mapping strategy for the 8 × 8 DB-MoMIMO system employing STMCM modulation scheme is tabulated in Table 3.7. The system schemes consist in [142]: (a) Encoder: ∗ Transmitted bit stream, {000001111110} . ∗ Transmitted symbol sequence, {0, 0, 2, 5, 1, 7} . ∗ Switching from even- to odd-numbered antennas after the third interval. It is noteworthy that to start the encoding, the P.S. is assumed to be 0 for the first time instant (starting with the even-indexed antennas) and the P.S. is assumed as 1 at the start of the fourth symbol interval to perform switching to odd-indexed antennas

136

3 Analysis of the Molecular Physical Layer’s Tasks

Table 3.7 Mapping strategy for selecting current antenna index based on previously activated antenna or previous state (P.S.) and current grouped input bits (i/p) in 8 × 8 MoMIMO system

I/P P.S. 0 1 2 3 4 5 6 7

00

01

11

10

Currently activated antenna index 2 4 6 0 3 5 7 1 4 6 0 2 5 7 1 3 6 0 2 4 7 1 3 5 0 2 4 6 1 3 5 7

(b) Decoder: ∗ First Stage. Let the tentative received sequence after choosing the maximum number of absorbed molecules for each symbol interval be {0, 0, 1, 5, 4, 7} . ∗ Second Stage. – Error detected in the third and fifth intervals due to parity mismatch. – Let the modified symbol sequence after replacing the erroneous symbols with the antenna corresponding to the second maximum count be {0, 0, 2, 5, 1, 7}. – The sequence is correct, so there is no need to repeat the process, and the final decoded antenna symbol sequence is {0, 0, 2, 5, 1, 7}. – At the end, each decoded symbol in the final sequence is demapped to two uncoded bits to obtain the final uncoded information bit stream. Because of the underlying mapping strategy, which successfully mitigates the negative effects of ISI and dominant ILI on the system error performance using STMCM MCD, the proposed coded modulation scheme in [142] has shown that for the majority of bit duration values, the proposed STMCM scheme performs better than the existing benchmark schemes (STMCM maximum count decoder). The mapping rules for the general NT × Nr DB-MoMIMO system can be envisioned as future work. • Example IX It is suggested in [143] that each sort of molecule can identify a layer, leading to the name layered molecular shift keying (LMoSK). In order to distinguish between layers, LMoSK employs a variety of molecules and transmits data at varied intervals inside each layer, designs two ML (maximum likelihood) detection techniques for LMoSK, ideal and practical, calculates the closed-form upper bound on the latter’s BER, and compares the BER performance of LMoSK with real-world ML detection to that of MoSK and MTPSK [143]. A DMC (diffusion-based molecular) system in a 3D unbounded environment with a point transmitter and a spherical passive receiver of volume V is

3.3 Molecular Communication Systems

137

considered. In [143], it is assumed that the transmitter can store Mtypes of information molecules and emit at most one type of them at a time. The receiver can detect and distinguish these arriving M-type molecules whose propagation properties can be characterized by Fick’s laws of diffusion with an identical diffusion coefficient D. Since the distance d between the transmitter and receiver is much greater than the radius r of the receiver, it can be assumed that the observed molecules in the spherical receiver are evenly distributed. In the proposed LMoSK, each type of molecule identifies a layer, and each symbol includes N time slots. Only a one-time slot of a symbol is chosen to emit the corresponding type of molecule in each layer. Assuming that each type of molecule emitted by the transmitter conveys b bits per emission and an LMoSK system with M types of a molecule can transmit B = Mb bits per symbol, then N = M + 2b − 1 [143]. After LMoSK modulation, the u-th sparse transmit symbol Su with M types of molecule emitted by the transmitter is given by Wen et al. [143]: .

Su = [S1 , S2 , . . . , SM ]T

(3.38)

where sm (m ∈ {1, 2, . . . , m}) with N × 1 dimensions represents the m-th layer and [·]T denotes the transpose of a matrix. Su is an M × N matrix where only one entry in each row is equal to one and no more than one entry in each column is equal to one. When the i-th (i ∈ {1, 2, . . . , N }) element of sm is equal to one and others are zero, the m-th type of molecule is emitted in the i-th time slot of Su [143]. At time t, after the transmitter releases a molecule based on Fick’s second law, the probability of a molecule arriving the center of the receiver can be  at d2 1 expressed as p(d, t) = (4π Dt) exp − 3/2 4Dt . By taking the derivative of the last mathematical expression with respect to time t and setting it equal to zero, it is possible to obtain the maximum molecular concentration at the receiver at the d2 peak time as td = 6D [143]. The time slot interval Te is chosen to be larger than td so that the receiver can detect the maximum molecular concentration of each emission. Hence, assuming that the transmitter and the receiver are in perfect synchronization, the receiver samples the molecular concentration at td after the time when molecules are emitted by the transmitter. The emitted molecules of L previous symbols may interfere with the current reception due to the nature of diffusion. Here, the set S = [S1 , . . . , Su , . . . , Sn ] corresponds to transmit signal of n symbols. Then, the molecular concentration of type m in the i-th time slot of symbol Su at the receiver after sampling is given by Wen et al. [143]: ym (u, i) =

.

Ni .   Epj + n(j Te + td ) S(m, w − j ) j =0

(3.39)

138

3 Analysis of the Molecular Physical Layer’s Tasks

where pj = p(d, j Te + td ), w = (u − 1)N + i, i ∈ [1, N ], S(m, w) ∈ S, Ni = min {LN, (u − 1)N} + i − 1, E is the number of releasing molecules of type m, and n(j Te + td ) denotes the counting noise [13] generated by a random process due to the transmission of molecules at time t = j Te + td [143]. For the modulation process in [143], it is assumed M types of molecule and each type of molecule convey b bits. Firstly, the source bits are divided into each symbol of B successive binary bits. Then, the bits of each symbol are divided into M layers as [b1 b2 . . . bM ], where bm is a b-bit binary vector representing bits of the m-th layer. Since each symbol has the same modulation procedure, one symbol is taken as an example to illustrate the details of the modulation as follows:

 (a) First, it takes an initial set A1 = 1, 2, . . . , 2b which consists of the whole possible emission time slots for the first layer and obtains

m1 by b1 . Then,  1, . . . , a1 , , , , , 2b . the m1 -th element a1 in the set A1 is found. Thus, ↑ m1 Subsequently, a1 represents the emission time slot of type

  molecules [143]. (b) A2 = A1 ∪ 2b + 1 \ {a1 } = 1, 2, . . . , 2b , 2b + 1 \ {a1 } is taken as the second set, and m2 is obtained by b2 of the second layer. Therefore, the m2 -th element a2 of the set A2 is chosen as the emission time slot for the molecules of type 2. The

next set A3 is selected by adding a new sequence of emission time slot 2b + 2 to A2 and removing {a2 } from it for the next layer’s modulation [143]. (c) Step 2 is repeated until the M-th layer is selected. Finally, the sequence of emission time slots of the symbol is given by {a1 , a2 , . . . , aM }, where ai represents the emission time slot of type molecules [143]. For the performance analysis, a theoretical upper bound on the BER of the practical ML receiver is derived. Then, it is assumed that Xt ∈ X is a transmit symbol while Xd ∈ X(Xd = Xt ) is an incorrectly detected symbol. The detected symbol is considered a symbol detection error that occurs when M  N Y (m, i)(X (m, i) − X (m, i)) ≤ 0 [143]. m=1 t d i=1 u For further analysis, it is assumed that Yu (m, i) follows Gaussian distribution, namely, Yu (m, i) ∼ N(∧u (m, i), ∧u (m, i)), due to a large number of emitted M  N Y (m, i)(X (m, i) − X (m, i)), and it also molecules E. Let Z = m=1 t d i=1 u obeys a Gaussian distribution, whose mean and variance are, respectively, given M  N ∧ (m, i)(X (m, i) − X (m, i)) and σ 2 =  M  N ∧ by μZ = m=1 t d Z m=1 i=1 u i=1 u (m, i)(Xt (m, i) − Xd (m, i))2 [143]. In the following, the probability that the transmit symbol Xt is improperly detected as Xd can be given by P (Z ≤ 0). Thus, the erroneous probability can be derived as [143]: .

P (Xt → Xs) = Q

 μZ  σZ

(3.40)

3.3 Molecular Communication Systems

139

∞ 2 where Q(x) = (2π )−1/2 x e−t /2 dt. Finally, the union bound technique an upper bound on the BER (with clearly better results) is obtained as [143]:

.

Pe ≤

1 2B B

2B 

2B 

t=1 d=1,d =t

P (Xt → Xd ) ∈ (Xt → Xd )

(3.41)

where e(Xt → Xd ) enumerates the number of error bits between Xt and Xd [143]. • Example X An innovative modulation technique is put forward in [144] to lower the ISI of MC by utilizing acids, bases, and salts. The time slot for delivering bit information “1” is further divided into three pieces by the authors. They introduce a strong base (hydroxide ions) to react with hydrogen ions and make sure that hydroxide ions can fully react with hydrogen ions, and they add a little salt to the solution to make sure that the next time slot transmits another bit of information and the solution remains neutral because the strong acid (hydrogen ions) in the current time slot may affect the next time slot. They discovered through simulation analysis that our suggested approach will significantly lower ISI and increase the accuracy of the information transfer, which may be used to construct a more effective macroscale molecular communication test bed. As shown in Fig. 3.30, in [144], the T x consists of three nozzles, which can emit strong acid (H + ), strong base (OH − ), and the salt (N aCl). The Rx is made of a pH sensor that can detect the pH value and decode the information according to the pH value. The acid or base should not have a very low or high pH value which could be destructive. The OH − can react with H + which will decrease the concentration of H + and ISI. The salt in this system has three advantages. Firstly,

Fig. 3.30 The new communication system

140

3 Analysis of the Molecular Physical Layer’s Tasks

the salt is neutral, and it almost cannot affect the pH of the solution. Secondly, due to the reaction of H + and OH − , it can produce water, and after a long time, the concentration of the solution will be lower and have a negative effect on the detection of the pH sensor. Thirdly, the salt can be a kind of information; as we know when we use MoSK modulation technique, the ISI will be immigrated [144]. Here, it is assumed that CH + (x, t) and COH −(x, t) represent the average spatiotemporal concentration of H + and OH − ions, respectively [144]. The information encoded can adjust the release of hydrogen ions only (i.e., strong acids), the release of hydroxide ions only (i.e., strong bases), or the release of salts only (i.e., NaCl). Note that if both hydrogen and hydroxide ions are released (i.e., strong acid and strong base), they will immediately combine and neutralize to form water molecules due to the very high forward reaction rate kf . Therefore, the current concentration of hydrogen ions and hydroxide ions in the solution is calculated as follows [144]:  T xreleases H

.

releases OH −

.

+

*

init (x) CH + (x, t = 0) = NH + δ(x) + CH + init (x) COH − (x, t = 0) = COH −

init (x) CH + (x, t = 0) = CH + init (x) COH − (x, t = 0) = NOH − δ(x) + C0H −

(3.42)

(3.43)

Here, it is assumed that the T x and the Rx maintain a high degree of synchronization at all times and the information molecules only collide with the surface of the Rx in a freely diffused channel. At this time, the diffusion motion of the information molecules released by the T x in the channel can be attributed to one-dimensional Brownian motion in the horizontal direction, thereby simplifying the analysis process [144]. When the sending nanomachine releases information molecules into the channel, after a period of time, these information molecules reach the Rx through diffusion and are absorbed by it. At the same time, the Rx will decode the received information molecules to obtain the original information, and finally, these information molecules received are removed from the channel. First, it is considered an acid pulse in which the number of acid molecules released is much greater than the concentration of ions in the channel. An approximate pulse response is obtained if it is assumed that there are no chemical interactions in the channel. If the velocity is constant and in the direction from the T x to the Rx and there is a sudden pulse response of x molar particles at time t = 0, then the function of the ion concentration received by the Rx over time is as follows [144]: .

C(x, t) =

√M 4π Dt

 2 exp − (x−vt) 4Dt

(3.44)

3.3 Molecular Communication Systems

141

where D is the diffusion coefficient of the particle, x is the diffusion distance between the transmitter and receiver, and v is the constant velocity [144]. To decrease the ISI, it is necessary to arrange the sequence of each information molecule properly. When we utilize hydrogen ions (H + ) to encode the bit information “1,” release salt (N aCl) to encode the bit information “0.” The information transmitted in each time slot is represented by an array, and the encoding by the TN can be formally defined as follows: * of the information 1, T N → H + a[i] = [144]. 0, T N → NaCl When the current time slot transmits bit information “1,” the T x releases strong acid (hydrogen ions); at the same time, when the next time slot transmits bit information “0” because there is a residual hydrogen ion concentration in the channel, it will cause an interference with transmitting bit information “0” under the current time slot. To solve this problem, it is possible to add a strong base (i.e., hydroxide ion) and ensure that the concentration of hydroxide ions is less than or equal to the concentration of hydrogen ions. Too many hydroxide ions will eliminate hydrogen ions, but it will affect the transmission of hydrogen ions under the next slot. Moreover, suppose the amount of hydroxide ions remains the same as the number of hydrogen ions because the release time of hydroxide ions and H + is different. In that case, a certain period is needed to ensure that hydroxide and hydrogen ions are fully neutralized. Then, we need to add N aCl instead of water. The reason for not adding water to the solution is that adding water will reduce the concentration of the solution, causing some errors in information transmission and affecting the accuracy of experimental results. If the amount of hydroxide ions is less than the number of hydrogen ions, ISI will be reduced but cannot be eliminated. Therefore, we need to add salt to replenish the easy concentration [144]. As shown in Fig. 3.31, two bits of information, “1” and “0,” are transmitted because the time slot for sending “0” or “1” is equal, which can be denoted as τ , and subsequently, the slot of sending “1” into three small time slots, i.e., τ1 , τ2 , and τ3 , is divided, making τ1 , τ2 , τ3 = τ . Therefore, when it is possible to transmit bit-“1,” at the start of the τ , T x emits strong acid (i.e., hydrogen ions); moles; after a while (τ1 ), T x then emits strong base (i.e., hydroxide ions); and a mole at the beginning of τ3 emits salt to make sure the hydroxide ions and hydrogen ions can react completely in this period. Finally, the concentration of the solution is 10−7 (i.e., the pH of the solution is 7), which is neutral and indicates that the ISI is well removed from the channel. When the bit information “0” was transmitted, to make the solution neutral, the T x can emit a little salt at the beginning of the time slot τ [144]. • Example XI In [145], it considered MISO (multiple-input single-output) topologies that incorporate nT x distinct spherical transmitters with radius rT x , whose centers are placed onto a uniform circular array (UCA) with the ability to emit molecules into the diffusion channel, as well as a spherical receiver with radius rRx that can absorb the molecules arriving at its surface therewithal record their azimuth

142

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.31 Time mechanism of releasing different ions when transmitting bit information “1” or “0”

and elevation angles. Note that the receiver’s center is perfectly placed onto the axis of the UCA with a distance of dRx from the center. The closest distance between the center of the UCA and any of the transmitters is dT x . (This topology is presented in Fig. 3.32a.) Additionally, the receiver sphere is partitioned into nT x different regions with the purpose that each region becomes a conjugate of the oppositely positioned transmitter and each Rx region transmitter conjugate is indexed from 0 to (nT x − 1), consecutively (region boundaries are further shown in Fig. 3.32b.). For [145] experiments, the values are selected as follows: nT x = 8, rT x = 0.5 μm, rRx = 5 μm, dRx = 15.5 μm, and dT x = 10 μm [145]. For the modulation in [145], the channel is used only once per symbol duration. In this way, information is encoded into the index of the intended T x antenna rather than any other aspect of the molecules, such as quantity, type, or temporal position. The packet size of the encoded information depends on the number of antennas in use, i.e., for nT x antennas, information is encoded into log2 (nT x )-bit long packets. Rx antenna collects the messenger molecules through its receiver regions for a defined symbol duration and detects the originating T x antenna. Information is then extracted through the index of the detected T x antenna. This general modulation scheme enables reliable information transmission with less channel use, directly implying less interference. In the proposed topology case, the receiver regions of Rx are compacted on the spherical centralized Rx surface. As aforementioned, molecules absorbed by Rx are recorded with their azimuth and elevation angles, enabling Rx to perfectly

3.3 Molecular Communication Systems

143

Fig. 3.32 Communication scenarios for nT x = 8 scenario. (a) MISO scenario in 3D view. (b) MISO scenario in 2D view. Black lines show the region boundaries on Rx, and each regiontransmitter conjugate is indexed from 0 to 7, consecutively

detect which receiver region the molecules hit during a symbol period. Absorbed molecules are processed to decode the information [145]. Besides, in [145], it proposed a machine learning-based approach that utilizes the time series data of absorbed molecule rates recorded by nT x regions of the receiver at each time instant. To train the proposed model, the considered communication scenario is simulated in which the time series of nT x regions according to the number of molecules arrived are recorded after a randomly selected transmitter is allowed to emit M molecules at time t = (k − 1)ts , where k ∈ {1, 2, . . . , w} , ts , and w denote the symbol window time and window number, respectively. For the experiments defined in [145], T = 5 s is split into w ∈ {3, 4, . . . , 10}, each of which corresponds to a symbol time of ts = T w . To simulate the random movement of the molecules in the driftless fluid environment, the total time of communication scenarios is divided into steps of length t = 10−4 s to perform computer simulations, and for each time interval [(k − 1)ts , kts ], molecules are emitted at instant time t = (k − 1)ts [145]. The position of each molecule is updated in the 3D space according to X(t +t) = X(t)+X; Y (t +t) = Y (t)+Y ; Z(t +t) = Z(t)+Z, where X, Y , and Z are independent and identically distributed Gaussian random numbers with μ = 0 and σ 2 = 2Dt, where the diffusion coefficient D is selected 79:4 μm2 s−1 . Furthermore, if a molecule hits the surface of the receiver, it is absorbed by the receiver and removed from the environment. Note that each receiver region records the number of absorbed molecules at every time instant, t = 10−4 s, which is then down-sampled to t = 10−1 s and normalized across regions to reduce computation complexity. Intuitively, the naive approach, namely, maximum count decoder (MCD), for predicting the correct transmitter

144

3 Analysis of the Molecular Physical Layer’s Tasks

for the kth window for one sample m can be formulated as [145]:

m Y . k

argmax i

kts  j =(k−1)ts

m Xi,j

(3.45)

m ), i = 1, . . . , n , andj = 1, . . . , T , i.e., X m is the set of where Xm = (Xi,j Tx t th the time series for the m sample containing molecule rates of nT x regions for each discrete time and is row-wise summed throughout the k th window. Thus, it can be found which region of the receiver is absorbing the maximum number of molecules, identical to the correct transmitter for the k th window and the mth sample denoted as Yˆkm , i.e., yˆ m = (Yˆkm ), k = 1, . . . , w [145]. The communication performance of the suggested MISO topology in Fig. 3.33 is evaluated through Monte Carlo simulations. For performance evaluations, a naive approach with Rx design of the maximum count decoder is formulated which is employed as the base modulation for comparisons. To compare the naive approach (maximum count decoder (MCD)) with an optimal Rx, a symbol-by-symbol maximum likelihood estimator (MLE) is also employed by considering the past L symbols to find the most likely symbol at each symbol

Fig. 3.33 Natural coding (NC) and gray coding (GC) bit error rate for maximum count decoder (MCD), machine learning (ML), symbol-by-symbol maximum likelihood estimation (MLE) approaches with tb = 0.166s

3.3 Molecular Communication Systems

145

Fig. 3.34 Natural coding (NC) and gray coding (GC) bit error rate for maximum count decoder (MCD), machine learning (ML), symbol-by-symbol maximum likelihood estimation (MLE) approaches with tb = 0.555s

duration. Communication simulations are conducted for both high and low data transmission rates. Also, a communication simulation with varying bit duration tb is performed to show the convergence of performance under the ISI and ILI effects [145]. The simulation’s output for the scenario with a high data transfer rate is shown in Fig. 3.34. When the receiver zones are arranged next to one another as segments of a spherical receiver, the likelihood of molecules being incorrectly received is very high. ILI is brought on as a result of large molecules being absorbed in nearby receiver regions. ISI and ILI become the main interference issues for molecular communications with large data transmission rates. As a result of substantial interference, MCD is prone to inaccuracy. The maximum likelihood estimator for each symbol, however, demonstrates the best results MCD can achieve. Due to its capacity to learn interference patterns more successfully than the optimal maximum likelihood estimation, the suggested method outperforms the optimal estimator in high data transmission rate scenarios [145]. Results from low data transmission rates show that ILI becomes the main cause of interference. Due to significant ILI, MCD is prone to mistakes. Yet, in terms of bit error rate performance, the suggested machine learning method greatly outperforms both MCD and the maximum likelihood estimation. Because

146

3 Analysis of the Molecular Physical Layer’s Tasks

ISI loses power in low data transmission rates, our suggested technique performs better at learning and identifying interference patterns [145]. • Example XII Multiple-input multiple-output (MIMO) technology, according to [146], enables molecular communication (MC) systems work better and increase communication throughput. Intersymbol interference (ISI) and interlink interference (ILI) difficulties still persist in molecular MIMO systems, despite index modulation (IM) techniques that help by storing the information in the antenna index. A stronger ability to battle ISI and ILI makes coding modulation techniques show promising benefits in terms of system error performance. Hence, a unique coded modulation system is proposed in [146] where the pulse position constellation is combined with the coded antenna indices. Due to its stronger ILI mitigation strength, the suggested approach is shown to produce superior error performance [146]. The system model under consideration in [146] assumes infinite planer transmitter (T x) and receiver (Rx) embedded with a finite number of point sources as transmit antennas and spherical absorbing receivers as receiver antennas. For synthetic MC scenarios, the physical protrusions from the T x and Rx aid in interacting with the external environment and are referred to as antennas in [146]. The transmitter and receiver nanodevices are attached to the cell membranes of the T x and Rx. It is assumed that only the antennas attached to Rx are capable of receiving the messenger molecules. The macroscale DBMC (diffusion-based molecular communication) environment is realized in [147]. The authors also consider aligned, symmetrical, and synchronized T x and Rx antennas in an unbounded three-dimensional diffusion without drift surroundings where molecules can freely propagate. The T x body is fully transparent to the messenger molecules, i.e., T x allows molecules to pass through it after transmission, and the Rx body is fully reflective, allowing the molecules to reflect from the surface after they hit. Hence, in [146], a uniform circular array (UCA) of T x and Rx antennas is considered where the number of transmit antennas (NT x ) and the number of receiver antennas (NRx ) equal to eight, i.e., NT x = NRx = 8. Consequently, an 8 × 8 molecular MIMO system is considered in Fig. 3.35. The channel for the considered scenario is simulated using Monte Carlo techniques. The mth channel coefficient is represented by hi,j [m] for the i th T x antenna and j th Rx antenna, where 0 ≤ i, j ≤ NT x − 1, 1 ≤ m ≤ P , and P denotes the ISI length corresponding to the channel memory. It is possible to approximate the number of received molecules at the j th Rx antenna and in the nth symbol duration, i.e., Rj [n], as a Gaussian random variable with sufficient accuracy and the probability of success as hi,j [m]. Letting N(·, ·) denote the normal distribution, then Rj [n] ∼ NT x−1 n N(μj [n], σj2 [n]), with μj [n] = k=n−P +1 i=0 si [k]hi,j [n − k + 1] and N n T x−1  si [k]hi,j [n − k + 1] × (1 − hi,j [n − k + 1]), where σj2 [n] = k=n−P +1 i=0

si [k] represents the number of transmitted molecules from the i th T x antenna point in the k th symbol duration [146].

3.3 Molecular Communication Systems

147

Fig. 3.35 The system model for 8 × 8 diffusion-based molecular MIMO system. The distance between the nearest points of T x and Rx antennas is denoted by dtr , and the distance between the projection of the spherical receiver antenna and the center of UCA is represented by dru . The radius of the spherical absorbing Rx antenna is given by rr

The proposed modulation analyzed in [146] establishes a PPM (Pulse Position Modulation)-based Molecular Spatio-Temporal Coded Modulation (MSTCM) SM (Spatial Modulation) scheme, i.e., Position Molecular Spatio-Temporal Coded Modulation (PMSTCM), where PPM is utilized along with space-time coded antenna indices to encode the information sequence. For the NTx × NRx DBMC (Diffusion Based Molecular Communication) system with NTx = NRx = δ. In [146], the uncoded information bit stream is first divided into groups– each containing B = log2 (γ ) + log2 (δ/2) bits, where the first log2 (γ ) bits in each grouped sequence, named as the γ -chip, represent the PPM symbol. The next uncoded log2 (δ/2) bits are coded to log2 (δ) bits using a mapping strategy that activates a particular antenna index depending on the previously activated antenna denoted by previous state (P .S), and then transmitting the γ -PPM symbol from that activated antenna. Consequently, the proposed (γ , δ)-

148

3 Analysis of the Molecular Physical Layer’s Tasks

PMSTCM scheme allows the transmission of coded information bit sequence x comprising of Bc = log2 (γ ) + log2 (δ) bits for each symbol duration. In [146] they mathematically represent the PMSTCM symbol, i.e., s[r], for the r th PPM Bc r symbol and r th activated antenna index as s[r] = k=B 2Bc r−k x[k] with c r−Bc +1 1 ≤ r ≤ Ld , where Ld denotes the length of the coded input symbol sequence and 0 ≤ x[k] ≤ γ δ − 1 denotes the k th bit in the coded information bit sequence [146]. Different modulation schemes for MC systems can be compared under two main constraints, i.e., the bit rate constraint and the molecular budget constraint. The bit rate constraint necessitates the same bit duration (Tb ) for different modulation schemes. In contrast, the molecular budget constraint mandates different schemes to be compared for an equal average number of transmitted molecules per bit duration denoted by MT /2 in [146]. The BER performance comparison of the proposed (γ , δ) PMSTCM scheme is presented using computer simulations. In [146], BER versus MT plots in Fig. 3.36 is provided. It can be observed from the figure that the increase in the molecular budget decreases the chances of error. For Tb = 0.6, it can be observed from Fig. 3.36 that higher-order PMSTCM schemes perform better than existing schemes [146]. To understand the BER performances deeply, Fig. 3.37 studies the variation of BER versus Tb for the considered system parameters. It can be seen from

Fig. 3.36 BER versus MT plots for the underlying 8 × 8 DB-MoMIMO system for the existing 2 and proposed modulation schemes; D = 79.4 μ ms , dtr = 4 μm, dru = 6.2 μm, rr = 4 μm, Tb = 0.6, input bit stream length = 60, and L = 6 [146]

3.3 Molecular Communication Systems

149

Fig. 3.37 BER versus Tb plots for the underlying 8 × 8 DB-MoMIMO system for the existing 2 and proposed modulation schemes; D = 79.4 μ ms , dtr = 4 μm, dru = 6.2 μm, rr = 4 μm, MT = 100, input bit stream length = 60, and L = 6 [146]

the figure that an increase in Tb leads to decay in ISI and enhanced ILI. For small values of Tb , most of the (γ , δ) PMSTCM schemes perform better than existing MPSM schemes. This is because the channel becomes ISI dominant for small Tb and system performance becomes more sensitive to ISI-caused errors. According to the mapping strategy, antennas activated in the current and previous symbol duration are not the same, which aids in combating dominant ISI arriving from the previous symbol duration. Consequently, unlike the MPSM scheme, the proposed PMSTCM scheme provides robustness against dominant ISI- and ILIcaused errors owing to the underlying MSTCM encoding strategy. Moreover, it is evident from Fig. 3.37 that lower-order PMSTCM scheme performs better than higher-order schemes for small values of Tb . The aforementioned behavior is attributed to the very small value of chip durations corresponding to the higher-order PPM constellations for a small range of Tb . The low chip durations increase ISI-caused errors and deteriorate the system performance of higherorder PMSTCM scheme for small Tb . As Tb increases, the chances of ILI increase, and consequently, higher-order PMSTCM schemes show better BER performance than existing modulation schemes due to their better ILI combating capability. The better ILI compatibility is attributed to the shorter chip slots, even for longer bit durations [146].

150

3 Analysis of the Molecular Physical Layer’s Tasks

3.3.6 Molecular Physical Layer Following the networking analysis, the molecular physical layer is separated into two sublayers, namely, the bio-nanomachine sublayer and the signaling sublayer, as indicated in Chap. 2 (Fig. 2.2). The primary properties and functions of these sublayers, as well as the physical transmission parameters of a molecular connection, are examined in this chapter. The biophysical foundation for addressing problems with the transmission, dissemination, and reception of information molecules is provided by the physical layer of a molecular communication architecture. The choice of hardware and interfaces suitable for molecular communication by bio-nanomachines is one problem. Using information molecules to represent a signal (signal modulation, discussed in Sect. 3.3.5), spreading information molecules in the environment (signal propagation, discussed in part 3.3.2), and boosting a signal that attenuates during propagation are additional problems (signal amplification). The capacity of molecular communication channels is one example of a performance factor that has problems [45]. Information theory, which is also used in computer networks, can be used to simulate how noise sources, such as the unpredictability in molecule propagation, affect the capacity of molecular communication channels. Several propagation models and modulation strategies have been used in the literature to examine molecular communication channels. There is still a need in this area for quantifying channel capacity using physically plausible models, comparing the capacities of various communication channels, and constructing molecular communication channels. Practically, molecular communication channels are not memoryless; thus, the channel capacity is different for different propagation models. Then, if .x n is a sequence of n successive transmission symbols and .y n is the corresponding received lim inf sup symbols, the channel capacity can be presented by .C = n→∞ x n I (Xn ; Y n ) [35, 45, 49, 78, 109, 148, 149]. On the other hand, because of signal attenuation and distortion, signal amplification is a crucial topic in MCs. Since information molecule concentration decreases with distance from the sender biological nanomachine, molecular communication depends on the movement of molecules in an aqueous environment. Between the sender and receiver bio-nanomachines, repeater bio-nanomachines that amplify information molecules can be positioned to solve this issue. To reach the receiver bio-nanomachine, for instance, intermediary repeater bio-nanomachines amplify the signal after the calcium ions have diffused from the sender bio-nanomachine. There aren’t many studies on signal amplification at the moment, and a number of problems still need to be resolved, like how to construct and use repeaters for other signal modulations besides calcium as well [35, 45, 49, 78, 109, 148, 149].

3.3 Molecular Communication Systems

3.3.6.1

151

Bio-nanomachine Sublayer

This sublayer of the physical layer abstracts physical details of bio-nanomachines and defines functionalities of bio-nanomachines [6].

3.3.6.1.1

Basic Components of Bio-nanomachines

A bio-nanomachine can have one or several of the following components: • Memory. It is a physical element that a biological machine utilizes to keep something in place. A bio-memory nanomachine may be represented by biochemical conditions in its internal environment, a three-dimensional arrangement of molecules in that environment, or by the three-dimensional arrangement of molecules in that environment [6]. • Biochemical conditions change as time progresses. The lifespan of biological molecules is finite, and memory-related molecules may degrade with time. As biochemical conditions change or when molecules that serve as memories degrade and stop serving as memories, memory loss at the bio-nanomachine sublayer happens [6]. • Molecule storage. A bio-nanomachine employs it as a physical part to store molecules. The environment for molecular communication known as “molecular storage” may be where molecules spread while they wait to be ingested by a biological machine. The inside environment of a bio-nanomachine may serve as a place for molecules to be stored. Molecule storage could be a physical element included into a bio-nanomachine, such as a vesicle or liposome that stores molecules [6]. • Biological molecules have a limited lifetime, and molecules stored in the molecule storage may deteriorate during their lifetime. Loss of molecules occurs when molecules in the molecule storage deteriorate and lose their functionality as the molecules. Loss of molecules also occurs when the molecular storage reaches its limit to store molecules [6]. • Process and processor. A process is a bio-manipulation nanomachine of molecules. A bio-nanomachine can modify the conformation of its own molecules, the molecules it stores, or the molecules in the environment to process molecules through biological processes. The physical element that a bio-nanomachine utilizes to process molecules is called a processor [6]. • Actuator. It is a physical element that a bio-nanomachine employs to move actively and spontaneously while using energy. A bio-actuator nanomachine could, for example, be a flagellum (a whiplike structure made of the protein flagellin) driven by a rotary motor, just like in bacteria. With the energy supplied from the ATP (adenosine triphosphate) hydrolysis, a bio-nanomachine may encapsulate its actuator, for example, in the form of leglike motor proteins present in molecular motors (such as kinesins) [6].

152

3 Analysis of the Molecular Physical Layer’s Tasks

• Identifier. It designates a single bio-nanomachine or a collection of them. A bionanomachine or a set of bio-nanomachines may be uniquely identified by one or more identifiers, or they may be located geographically [6]. – * Physical identifier. A bio-nanomachine has a physical identity connected to it (or a group of bio-nanomachines). It uniquely identifies a bio-nanomachine (also known as an individual identifier) or a collection of bio-nanomachines (also known as a group identifier) that have one or more shared traits (e.g., bio-nanomachines that provide the same functionality, such as carrying the same type of drug molecules or sensing the same type of molecules in the environment). A particular kind of molecule connected to a bio-nanomachine could serve as a physical identifier [6]. – * Location identifier. In the molecular communication environment, a location identifier uniquely identifies a place (i.e., an individual identifier) or many locations (i.e., a group identifier). Create an addressable space in the molecular communication environment based on concentration gradients of “guide” molecules; a location in the space is distinguished (and hence identifiable) by a concentration degree of guide molecules. This is one method of implementing a location identifier. A location identifier can be a particular kind of molecule associated with a site or a collection of locations, much like the physical identifier [6].

3.3.6.1.2

Bio-nanomachine Sublayer Functionalities

A bio-nanomachine implements a set of simple functionalities including lifesustaining, actuation, and/or molecule processing functionalities [6]. • Life-sustaining functionalities. (a) Acquire and expend energy. Bio-nanomachines acquire and expend energy to perform their functionalities [6]. (b) Replicate. Bio-nanomachines make a copy of themselves and produce bionanomachines capable of the same or similar functionality [6]. (c) Terminate functioning and decompose. Bio-nanomachines have a finite lifespan, and over that time, their functionality may decline. When they become inoperable, bio-nanomachines remove themselves from the environment [6]. • Actuation functionality. – Move. Bio-nanomachines are mobile in their surroundings. Bio-nanomachines are capable of passive movement, such as following the fluids in the environment or stochastic diffusion in the environment. Moreover, bio-nanomachines may move actively in a certain direction [6, 147]. – Molecule processing functionalities.

3.3 Molecular Communication Systems

153

(a) Capture molecules. Bio-nanomachines capture molecules directly from another bio-nanomachine or from the environment [6]. (b) Capture molecules. Bio-nanomachines capture molecules or from the environment [6]. (c) Store molecules. Molecule storage is where bio-nanomachines keep the molecules they capture or make. In order to store molecules, bio-nanomachines may employ either their internal environment or the environment for molecular communication [6]. (d) Release molecules. Bio-nanomachines release molecules into the environment or directly into other bio-nanomachines [6]. (e) Synthesize molecules. By biochemical processes, bio-nanomachines create molecules from molecules that already exist in their internal and external surroundings. At the signaling sublayer, for example, new sorts of molecules could be created as signal molecules or molecules with information encoded on them [6]. (f) Detect molecules (and biochemical conditions). A bio-nanomachine can identify a certain class of molecule in the surrounding environment. Its capacity may be expanded by bio-nanomachines to detect biochemical conditions in the environment, such as the quantity or concentration of particular molecules [6]. (g) Modify molecules (and biochemical conditions). In their internal and external settings, bio-nanomachines alter their own features or the characteristics of molecules (such as three-dimensional structure or conformation) [6]. (h) Remember the state. Bio-nanomachines can recall their current condition. Moreover, bio-nanomachines may have memory in the form of the quantity of molecules they have stored inside of them [6]. (i) Change the state. Bio-nanomachines switch between straightforward states. A bio-nanomachine can alter its state through conformational changes, whether it is in the state where its three-dimensional structure is functional or nonfunctional. If a bio-internal nanomachine’s storage of molecules accurately reflects its state, it may be possible for it to change its state by altering that number [6]. (j) Keep track of time. A clock may be biochemically implemented by bionanomachines. The biological clock that naturally occurs in bio-nanomachines, such as a circadian clock, may also be used. (k) Self-feedback. When a bio-nanomachine feeds back its own output, it may regulate itself. Self-feedback in bio-nanomachines may be used to build oscillatory chemical clocks for timekeeping [6].

3.3.6.2

Signaling Sublayer

A bio-nanomachine can communicate with other bio-nanomachines using the signaling sublayer of the physical layer. Biomaterials are used for signaling at the signaling sublayer [6].

154

3 Analysis of the Molecular Physical Layer’s Tasks

Fig. 3.38 Signaling sublayer in molecular communication based on Shannon’s model of communication

3.3.6.2.1

Signaling Sublayer Communication Model.

According to Shannon’s [150–153] model of communication, basic components of the signaling sublayer are the following (Fig. 3.38): (a) Message. It is a notion, knowledge, communication, or assertion that both the information source and the information destination comprehend [6]. (b) Information source. A message or series of messages are generated and sent to the information destination. At the application layer, the information source could be a person, a bio-machine, or a standard device [6]. (c) Transmitter. In order to create a signal (information-encoded molecules, also known as signal molecules) that may be transmitted via the molecular communication channel, it performs an operation on the message. In the molecular communication environment, the transmitter could be either a bio-nanomachine or a conventional device (in the external environment) [6]. (d) Molecular communication channel. It is the medium used to transmit the signal (e.g., signal molecules) between directly communicating transmitter and receiver. A molecular communication channel may be implemented such that it inherently has broadcast nature (a broadcast channel). For instance, it may be the space in the environment where signal molecules diffuse and propagate stochastically from a transmitter to a receiver (or receivers). A molecular communication channel may also be implemented such that it inherently has a point-to-point nature (a point-to-point channel). For instance, it may be a protein filament along which protein motors carry signal molecules from a transmitter to a receiver [6]. (e) Receiver. It extracts the message from the signal by performing the opposite action of what the transmitter did (e.g., signal molecules). In the molecular

3.3 Molecular Communication Systems

155

communication environment, the receiver could be either a bio-nanomachine or a regular device (in the external environment) [6]. (f) Information destination. It is the entity to which the message is addressed. A person, a standard device, or a bio-nanomachine could be the information destination at the application layer [6]. (g) Signal molecule storage. It is a material that holds signal molecules. Similar to the bio-nanomachine sublayer, signal molecule storage may take the form of a physical entity embedded in a bio-nanomachine, such as a vesicle or liposome, which is embedded in the bio-nanomachine to store molecules, or it may take the form of a molecular communication environment where the signal molecules diffuse and wait for the bio-nanomachine to absorb them. Signal molecule loss happens when signal molecules in the storage for signal molecules degrade and stop functioning as the signal molecule. When the capacity of the molecular signal storage to hold signal molecules is reached, signal molecules are also lost [6]. (h) Noise source. In a molecular communication route, it results in a mistake or unwanted random disturbance of a valuable signal (such as signal molecules). Environmental noise from the molecular communication environment and the signal from other communications both contribute to noise at the signaling sublayer (cross talk) [60]. Sources of biochemical, thermal, and physical noise are all environmental noise sources (as discussed previously). Signals (like signal molecules) transferred through a different communication channel might produce cross talk. Cross talk may happen, for instance, when a transmitter and a receiver speak to one another by employing signal molecules that are diffusing in the surrounding air. A transmitter’s signal molecules spread out in the environment, float in the direction of another transmitter/receiver pair, and then begin to obstruct communication. Cross talk can also happen, for example, when a transmitter and a receiver interact using protein motors that move along a protein filament that connects the transmitter and the receiver and convey signal molecules. In order to communicate with another transmitter/receiver pair, protein filaments carrying signal molecules may separate from one and disseminate in the environment before attaching to another filament containing the same proteins signaling sublayer.

3.3.6.2.2

Signaling Sublayer Functionalities

(a) Signal modulation. The transmitter’s ability to change the signal’s characteristics, such as its signal molecules, in order to represent a message to send is one of its functionalities. Selecting one type of molecule from a group of identifiable types of molecules, each type conveying a specific piece of information, is one method for modulating signals (i.e., molecule shift keying (MoSK)). Another method involves using a single type of (identifiable) molecule and altering the patterns of its release, for example, adjusting the number of molecules to communicate (using amplitude shift keying (ASK)) and the interval at

156

3 Analysis of the Molecular Physical Layer’s Tasks

which the molecules are transmitted (i.e., frequency shift keying (FSK)). Also known as nucleotide shift keying, information can be encoded into DNA base sequences (NSK) [87]. Functionalities at the bio-nanomachine sublayer like “release molecules” in a regulated manner, “synthesize molecules,” and “modify molecules” may facilitate signal modulation. The bio-nanomachine sublayer receives information from the signaling sublayer about the types of molecules to release and their patterns of release [6]. In Section 3.3.5, modulation techniques were widely explained. (b) Signal demodulation. The ability to reverse signal modulation is a capability of the receiver. Signal demodulation at the bio-nanomachine sublayer may be supported by the functions of molecules involved in processing. As was previously said, bio-nanomachines may “detect molecules” to detect both the concentration of molecules in the environment and the existence of a certain type of molecules in the environment (if the signal modulation method is based on the type of molecules and if the signal modulation mechanism is based on changing the number of molecules to transmit). The signaling sublayer receives information from the bio-nanomachine sublayer on the types of molecules collected and their patterns [6]. (c) Signal molecule transmission/reception. It is the capability of both the transmitter and the receiver to transmit signal molecules onto the molecular communication channel (“signal transmission”) and to receive incoming signal molecules (“signal reception”), respectively. Functionalities like “release molecules” in a regulated manner and “catch molecules” may enhance signal transmission/reception at the bio-nanomachine sublayer. The signaling sublayer informs the bio-nanomachine sublayer about the sort of molecules to employ as signaling molecules and the timing of molecule release while conveying signal molecules. The signaling sublayer receives signal molecules, and the bionanomachine sublayer transmits information on the nature and timing of the molecules’ reception [6]. (d) Signal propagation. It is the capability of both the transmitter and the receiver to transmit signal molecules onto the molecular communication channel (“signal transmission”) and to receive incoming signal molecules (“signal reception”), respectively. Functionalities like “release molecules” in a regulated manner and “catch molecules” may enhance signal transmission/reception at the bionanomachine sublayer. The signaling sublayer informs the bio-nanomachine sublayer about the sort of molecules to employ as signaling molecules and the timing of molecule release while conveying signal molecules. The signaling sublayer receives signal molecules, and the bio-nanomachine sublayer transmits information on the nature and timing of the molecules’ reception [6]. The transmission medium may aid in the signal propagation necessary for molecules to move around in a molecular communication environment. Active or passive signal propagation techniques may be used on the communication channel. A molecular communication channel can be either a point-to-point or a broadcast channel. Although there are theoretically four possible combinations of the molecular communication channel type and the

3.3 Molecular Communication Systems

157

mode of propagation, passive propagation over a point-to-point channel may not be practical because propagating directionally over a point-to-point channel requires active propagation, which uses energy. The implementation of passive signal propagation across a broadcast channel requires little to no effort because it makes use of the available space or fluid in the surroundings. Establishing connectivity between bio-nanomachines, such as a concentration gradient that self-propelling organisms follow and a network of protein filaments that protein motors move along, is necessary for active signal propagation (across either a broadcast channel or a point-to-point channel). Such connectedness must be created beforehand, either naturally or through the use of autonomous activities (e.g., dynamic instability of protein filaments) [6]. The bio-nanomachine sublayer may not receive instructions from the signaling sublayer about the best method of propagation to use. This is due to the possibility that molecular communication systems are a priori intended to use either passive or active modes of propagation rather than giving users the choice to switch between them as necessary [6, 154]. (e) Signal relay. For long-distance signal propagation, it is a functionality to propagate and amplify signals (e.g., signal molecules). Relay or repeater bionanomachines that amplify the signal, for example, by raising the concentration of signal molecules, can be introduced into the environment as one method of a signal relay. At a signal relay bio-nanomachine, functionalities such as “capture molecules” (to capture incoming signal molecules), “synthesize molecules,” and “release molecules” (to create new signal molecules and release newly created signal molecules into the environment) may support a signal relay (to release signal molecules stored in internal molecule storage into the environment). The signaling sublayer informs the bio-nanomachine sublayer at a relay machine about the level of amplification [6]. (f) Signal multiplexing. It is a capability to send many aggregate signals across a common molecular communication channel. Time division multiplexing (TDM) is a method of signal multiplexing in which several bio-nanomachines (or groups of bio-nanomachines) broadcast signal molecules at various times. Bio-nanomachine functions of “keep track of time” and “release molecules” in a regulated manner may facilitate TDM-based signal multiplexing. The bionanomachine sublayer receives information from the signaling sublayer on when to release chemicals [6]. (g) Signal molecule error handling. This functionality allows for the detection and potential correction of signaling sublayer noise-related problems. Incorporating channel codes, such as Hamming codes, into a pattern of molecules that convey data is one method of error handling (or error correction). Functionalities like “synthesize molecules” and “release molecules” in a controlled manner may help signal molecule error management at the bio-nanomachine sublayer. The bio-nanomachine sublayer receives information from the signaling sublayer about the types of molecules to release and their patterns of release [6].

158

3 Analysis of the Molecular Physical Layer’s Tasks

(h) Addressing. The capability to designate a receiver (or a group of receivers) that receives signal molecules or a location (or a number of locations) to which molecular frames are sent is found at the signaling sublayer. A physical address (i.e., an address that identifies a receiver or a group of receivers) and a location address may both be supported by addressing at the signaling sublayer (i.e., an address that identifies a location or multiple locations). It may also support a group address (i.e., an address that identifies a group of receivers or multiple locations) as well as an individual address (i.e., an address that designates a single recipient or a single place) [6]. * Physical address. A receiver (i.e., an individual address) or a group of receivers that share the same characteristic or characteristics are uniquely identified by their physical addresses. For example, receivers that perform the same function, such as transporting the same kind of drug molecules or sensing the same kind of molecules in the environment, are addressed physically (i.e., a group address). A bio-nanomachine physical identifier and “capture molecules” and “detect molecules” functions at the bionanomachine sublayer may enable a physical address at the signaling sublayer [6]. * Location address. In the molecular communication environment, a location address uniquely identifies a place (i.e., an individual address) or a collection of locations (i.e., a group address). Similar to a physical address, a signaling sublayer location address may be supported by a location identifier and functions like “capture molecules” and “detect molecules” at the bionanomachine sublayer. If the bio-nanomachine sublayer uses molecules to implement a location identifier, then molecules that bind to the location identifier molecules at the bio-nanomachine sublayer may be used as a location address at the signaling sublayer, just like the physical address. The bio-nanomachine sublayer receives information from the signaling sublayer regarding the address, the type of address it employs (physical address, location address, individual address, group address), and group membership (if it is a group address) [6]. (i) Storing signal molecules. Signal molecules may be kept in the corresponding signal molecule storage of the transmitter and receiver [6]. (j) Feedback. It is a process where information from the past or present affects an event that will happen in the future or present. It is possible for the signaling sublayer to provide feedback to itself or for there to be feedback between the transmitter and the receiver. At the bio-nanomachine sublayer, “self-feedback” may assist self-feedback. Information does not have to be passed from the signaling sublayer to the bio-nanomachine sublayer. It might merely activate the “self-feedback” at the sublayer of the bio-nanomachine [6].

References

159

References 1. M. Egan, M. Kuscu, M.T. Barros, M. Booth, A. Llopis-Lorente, M. Magarini, D.P. Martins, M. Schäfer, P. Stano, Toward interdisciplinary synergies in molecular communications: perspectives from synthetic biology, nanotechnology, communications engineering and philosophy of science. Life 13(1), 208 (2023) 2. A. Awad, M.M. Fouda, M.M. Khashaba, E.R. Mohamed, K.M. Hosny, Utilization of mobile edge computing on the Internet of Medical Things: A survey. ICT Express, Elsevier (2022) 3. D. Chen, A.G. Strang, A.W. Eckford, P.J. Thomas, Explicitly solvable continuous-time inference for partially observed Markov processes (2023). Preprint arXiv:2301.00843 4. X. Chen, M. Wen, F. Ji, Y. Huang, Y. Tang, A.W. Eckford, Detection interval for diffusion molecular communication: How long is enough? (2022) Preprint arXiv:2204.08636 5. M. Rezaei, H.K. Rudsari, M. Javan, N. Mokari, E.A. Jorswieck, M. Orooji, Molecular communication transmitter design in limited-capacity storage regime. IEEE Trans. NanoBiosci. 22, 212–222 (2022) 6. T. Nakano, T. Suda, Y. Okaie, M.J. Moore, A.V. Vasilakos, Molecular communication among biological nanomachines: A layered architecture and research issues. IEEE Trans. Nanobiosci. 13(3), 169–197 (2014) 7. H.K. Rudsari, M. Zoofaghari, M. Veletic, J. Bergsland, I. Balasingham, The end-to-end molecular communication model of extracellular vesicle-based drug delivery (2022). Preprint arXiv:2207.01875 8. S. Senturk, I. Kok, F. Senturk, Internet of nano, bio-nano, biodegradable and ingestible things: A survey (2022). Preprint arXiv:2202.12409 9. S. Lotter, L. Brand, V. Jamali, M. Schäfer, H.M. Loos, H. Unterweger, S. Greiner, J. Kirchner, C. Alexiou, D. Drummer et al., Experimental research in synthetic molecular communications–part I: Overview and short-range systems (2023). Preprint arXiv:2301.06417 10. S. Lotter, L. Brand, V. Jamali, M. Schäfer, H.M. Loos, H. Unterweger, S. Greiner, J. Kirchner, C. Alexiou, D. Drummer, G. Fischer, A. Buettner, R. Schober, Experimental research in synthetic molecular communications – part II: Long-range communication (2023) 11. Y. Li, L. Lin, W. Guo, D. Zhang, K. Yang, Error performance and mutual information for iont interface system. IEEE Int. Things J. 9, 9831–9842 (2022) 12. O. Yetimoglu, M.K. Avci, B.C. Akdeniz, H.B. Yilmaz, A.E. Pusane, T. Tugcu, Multiple transmitter localization via single receiver in 3-d molecular communication via diffusion. Digital Signal Proc. 124, 103185 (2022) 13. Z.P. Li, J. Zhang, T.C. Zhang, Concentration aware routing protocol in molecular communication nanonetworks, in Applied Mechanics and Materials, vol. 556 (Trans Tech Publications, Wollerau, 2014), pp. 5024–5027 14. L.F. Borges, M.T. Barros, M. Nogueira, Toward reliable intra-body molecular communication: an error control perspective. IEEE Commun. Mag. 59(5), 114–120 (2021) 15. R. Paridar, N. Mokari, E. Jorswieck, M.R. Javan, On the reception process of molecular communication-based drug delivery (2021). Preprint arXiv:2112.12485 16. T. Islam, E. Shitiri, H.-S. Cho, In-body sequential multi-drug delivery scheme using molecular communication. IEEE Access 10, 39975–39985 (2022) 17. F. Büther, F.-L. Lau, M. Stelzner, S. Ebers, A formal definition for nanorobots and nanonetworks, in Internet of Things, Smart Spaces, and Next Generation Networks and Systems (Springer, Berlin, 2017), pp. 214–226 18. I.F. Akyildiz, F. Brunetti, C. Blázquez, Nanonetworks: a new communication paradigm. Comput. Netw. 52(12), 2260–2279 (2008) 19. B. Atakan, S. Galmes, Effects of framing errors on the performance of molecular communications with memory. IEEE Access 8, 19970–19981 (2020)

160

3 Analysis of the Molecular Physical Layer’s Tasks

20. L. Hong, W. Chen, F. Liu, Cooperative molecular communication for nanonetwork, in 2014 Sixth International Conference on Ubiquitous and Future Networks (ICUFN) (IEEE, Piscataway, 2014), pp. 369–370 21. S.F. Bush, Interoperable nanoscale communication [future directions]. IEEE Consum. Electron. Mag. 6(2), 39–47 (2017) 22. S.F. Bush, J.L. Paluh, G. Piro, V. Rao, R.V. Prasad, A. Eckford, Defining communication at the bottom. IEEE Trans. Mol. Biol. Multi-Scale Commun. 1(1), 90–96 (2015) 23. A. Nayyar, V. Puri, D.-N. Le, Internet of nano things (iont): next evolutionary step in nanotechnology. Nanosci. Nanotechnol. 7(1), 4–8 (2017) 24. P. Kaliraj, T. Devi, Securing IoT in Industry 4.0 Applications with Blockchain (CRC Press, Boca Raton, 2021). [Online]. Available: https://books.google.com.ec/books?id= 771FEAAAQBAJ 25. A.J. Jani, Estimating nodes’ number in a nanonetwork using two algorithms, in Intelligent Sustainable Systems (Springer, Berlin, 2023), pp. 645–652 26. T. Khan, M. Civas, O. Cetinkaya, N.A. Abbasi, O.B. Akan, Nanosensor networks for smart health care, in Nanosensors for Smart Cities (Elsevier, Amsterdam, 2020), pp. 387–403 27. S. Fancher, A. Mugler, Fundamental limits to collective concentration sensing in cell populations. Phys. Rev. Lett. 118(7), 078101 (2017) 28. L. Wu, Y. Shen, M. Li, F.-X. Wu, Network output controllability-based method for drug target identification. IEEE Trans. Nanobiosci. 14(2), 184–191 (2015) 29. D. Bi, Y. Deng, Spatiotemporal control of genetic circuit with pulse generation for molecular communication, in, IEEE Global Commun. Confer. IEEE 2021, 1–6 (2021) 30. IEEE recommended practice for nanoscale and molecular communication framework. IEEE Std 1906.1-2015 (2016), pp. 1–64 31. IEEE standard data model for nanoscale communication systems. IEEE Std 1906.1.1-2020 (2020), pp. 1–142 32. T. Nakano, A networking paradigm inspired by cell communication mechanisms, in Nanotechnology: Concepts, Methodologies, Tools, and Applications (IGI Global, Pennsylvania, 2014), pp. 21–30 33. D.B. Menendez, V.R. Senthivel, M. Isalan, Sender-receiver systems and applying information theory for quantitative synthetic biology. Current Opin. Biotechnol. 31, 101–107 (2015) 34. I. Iqbal, M. Nazir, A. Sabah, Design of energy-efficient protocol stack for nanocommunication using greedy algorithms. J. Comput. Netw. Commun. 2022 (2022) 35. M.J. Salariseddigh, U. Pereg, H. Boche, C. Deppe, V. Jamali, R. Schober, Deterministic identification for molecular communications over the Poisson channel (2022). Preprint arXiv:2203.02784 36. A.D. Odysseos, T. Nakano, Bionanomachine diagnostics and nanonetwork therapeutic in brain malignancies with bionanodevice interfaces. IEEE Trans. Mol. Biol. Multi-Scale Commun. 8, 28–35 (2021) 37. J.T. Gómez, K. Pitke, L. Stratmann, F. Dressler, Age of information in molecular communication channels. Digital Signal Process. 124, 103108 (2022) 38. S.S. Somathilaka, D.P. Martins, S. Balasubramaniam, Information flow of cascading bacterial molecular communication systems with cooperative amplification, in IEE ICC 2022 (IEEE, Piscataway, 2022) 39. W. Pan, X. Chen, X. Yang, N. Zhao, L. Meng, F.H. Shah, A molecular communication platform based on body area nanonetwork. Nanomaterials 12(4), 722 (2022) 40. Y. Sun, H. Bian, Y. Chen, A photolysis-assist molecular communication for tumor biosensing. Sensors 22(7), 2495 (2022) 41. L. Chouhan, M.-S. Alouini, Rescaled Brownian motion of molecules and devices in threedimensional multi-user mobile molecular communication systems. IEEE Trans. Wirel. Commun. 21, 10472–10488 (2022) 42. T. Nakano, M. Moore, A. Enomoto, T. Suda, Part II Recent research and development in molecular communication technology. J. Natl. Instit. Informat. Commun. Technol. 55(4) (2008)

References

161

43. T. Nakano, Biologically inspired network systems: a review and future prospects. IEEE Trans. Syst. Man Cybern. Part C (Appl. Rev.) 41(5), 630–643 (2010) 44. T. Furubayashi, Y. Sakatani, T. Nakano, A. Eckford, N. Ichihashi, Design and wet-laboratory implementation of reliable end-to-end molecular communication. Wirel. Netw. 24(5), 1809– 1819 (2018) 45. T. Nakano, A.W. Eckford, T. Haraguchi, Molecular Communication (Cambridge University Press, Cambridge, 2013) 46. M. Dong, W. Li, X. Xu, Evaluation and modeling of HIV based on communication theory in biological systems. Infect. Genet. Evolut. 46, 241–247 (2016) 47. F. Walsh, S. Balasubramaniam, D. Botvich, T. Suda, T. Nakano, S.F. Bush, M. Foghlú, Hybrid dna and enzyme based computing for address encoding, link switching and error correction in molecular communication, in International Conference on Nano-Networks (Springer, Berlin, 2008), pp. 28–38 48. E.J. Alarcón Cot, R. Cid Fuentes, A. Davy, L. Felicetti, M. Femminella, P. Lio, G. Reali, J. Solé Pareta, Molcomml: the molecular communication markup language, in Proceedings of the 3rd ACM International Conference on Nanoscale Computing and Communication (ACM NANOCOM 2016): New York, USA: September 28–30, 2016 (Association for Computing Machinery (ACM), New York, 2016) 49. T. Nakano, M.J. Moore, F. Wei, A.V. Vasilakos, J. Shuai, Molecular communication and networking: opportunities and challenges. IEEE Trans. Nanobiosci. 11(2), 135–148 (2012) 50. M. Kuscu, O.B. Akan, Detection in molecular communications with ligand receptors under molecular interference. Digital Signal Process. 124, 103186 (2022) 51. T. Nakano, Y. Okaie, T. Hara, Applications of Molecular Communication Systems (Springer International Publishing, Cham, 2020), pp. 32–37 52. S. Ahmed, J. Hu, S.M. Naqvi, Y. Zhang, L. Linze, A.M. Iderawumi, Molecular communication network and its applications in crop sciences. Planta 255(6), 1–12 (2022) 53. Z. Cheng, J. Sun, J. Yan, Y. Tu, Optimizations for mobile MIMO relay molecular communication via diffusion with network coding. KSII Trans. Int. Informat. Syst. 16(4), 1373–1391 (2022) 54. I. Isik, How mobility of transmitter and receiver affects the communication quality. AIP Adv. 12(2), 025205 (2022) 55. A. El-atty, M. Saied, R. Bidar, E.-S.M. El-Rabaie, Embedded nano relay for intra-body network-based molecular communications. Wirel. Personal Commun. 125, 1–18 (2022) 56. X. Liao, Z. Jia, Z. Wu, B. Liu, X. Wang, X. Jiang, Modulation and signal detection for diffusive-drift molecular communication with a mobile receiver. Mob. Informat. Syst. 2021, 1–17 (2021) 57. S. Ghavami, R. Adve, F. Lahouti, Information rate analysis of ask-based molecular communication systems with feedback. Nano Commun. Netw. 28, 100339 (2021) 58. Z. Cheng, J. Yan, J. Sun, Y. Tu, K. Chi, Joint optimizations of relays locations and decision threshold for multi-hop diffusive mobile molecular communication with drift. IEEE Trans. NanoBiosci. 21, 454–465 (2022) 59. N. Farsad, Molecular communication: Interconnecting tiny nanobio devices. GetMob. Mob. Comput. Commun. 22(2), 5–10 (2018) 60. M. Civas, A. Abdali, M. Kuscu, O.B. Akan, Frequency-domain detection for molecular communications (2023). Preprint arXiv:2301.01049 61. S. Balasubramaniam, S. Ben-Yehuda, S. Pautot, A. Jesorka, Y. Koucheryavy et al., A review of experimental opportunities for molecular communication. Nano Commun. Netw. 4(2), 43– 52 (2013) 62. S. Sarkar, S. Choubey, Efficacy of information transmission in cellular communication (2020). Preprint arXiv:2010.05291 63. D. Bi, A. Almpanis, A. Noel, Y. Deng, R. Schober, A survey of molecular communication in cell biology: Establishing a new hierarchy for interdisciplinary applications (2020). arXiv e-prints, pp. arXiv–2009

162

3 Analysis of the Molecular Physical Layer’s Tasks

64. S. Figueiredo, N. Souto, F. Cercas, Low-complexity channel codes for reliable molecular communication via diffusion. Sensors 22(1), 41 (2021) 65. P. Hofmann, J.A. Cabrera, R. Bassoli, M. Reisslein, F.H.P. Fitzek, Coding in diffusion-based molecular nanonetworks: A comprehensive survey. IEEE Access 11, 16411–16465 (2023) 66. Z. Wei, B. Li, W. Hu, W. Guo, C. Zhao, Hamming-Luby rateless codes for molecular erasure channels. Nano Commun. Netw. 23, 100280 (2020) 67. J. Yu, H. Li, Convolutional codec implemented by genetic circuits for molecular communication. IEEE Trans. NanoBiosci. 22, 78–91 (2022) 68. M.T. Barros, S. Balasubramaniam, B. Jennings, Comparative end-to-end analysis of ca 2+-signaling-based molecular communication in biological tissues. IEEE Trans. Commun. 63(12), 5128–5142 (2015) 69. G. Yu, M. Yi, Y. Jia, J. Tang, A constructive role of internal noise on coherence resonance induced by external noise in a calcium oscillation system. Chaos Solitons Fractals 41(1), 273–283 (2009) 70. X. Huang, Y. Fang, N. Yang, A survey on estimation schemes in molecular communications. Digital Signal Process. 124, 103163 (2022) 71. A. Ahmadzadeh, H. Arjmandi, A. Burkovski, R. Schober, Comprehensive reactive receiver modeling for diffusive molecular communication systems: reversible binding, molecule degradation, and finite number of receptors. IEEE Trans. Nanobiosci. 15(7), 713–727 (2016) 72. N.V. Sabu, A.K. Gupta, Analysis of diffusion based molecular communication with multiple transmitters having individual random information bits. IEEE Trans. Molec. Biol. Multi-Scale Commun. 5(3), 176–188 (2019) 73. F. Gulec, B. Atakan, Localization of a passive molecular transmitter with a sensor network, in International Conference on Bio-inspired Information and Communication Technologies (Springer, Berlin, 2020), pp. 317–335 74. C.T. Chou, Impact of receiver reaction mechanisms on the performance of molecular communication networks. IEEE Trans. Nanotechnol. 14(2), 304–317 (2015) 75. S. Mohamed, D. Jian, L. Hongwei, Z. Decheng, Molecular communication via diffusion with spherical receiver & transmitter and trapezoidal container. Microprocess. Microsyst. 74, 103017 (2020) 76. M. Pierobon, I.F. Akyildiz, A physical end-to-end model for molecular communication in nanonetworks. IEEE J. Sel. Areas Commun. 28(4), 602–611 (2010) 77. M. Kuscu, O.B. Akan, On the physical design of molecular communication receiver based on nanoscale biosensors. IEEE Sensors J. 16(8), 2228–2243 (2016) 78. N. Farsad, H.B. Yilmaz, A. Eckford, C.-B. Chae, W. Guo, A comprehensive survey of recent advancements in molecular communication. IEEE Commun. Surv. Tutor. 18(3), 1887–1919 (2016) 79. H. Awan, C.T. Chou, Molecular communications with molecular circuit-based transmitters and receivers. IEEE Trans. Nanobiosci. 18(2), 146–155 (2019) 80. V. Jamali, A. Ahmadzadeh, N. Farsad, R. Schober, Constant-composition codes for maximum likelihood detection without CSI in diffusive molecular communications. IEEE Trans. Commun. 66(5), 1981–1995 (2018) 81. W. Guo, T. Asyhari, N. Farsad, H.B. Yilmaz, B. Li, A. Eckford, C.-B. Chae, Molecular communications: channel model and physical layer techniques. IEEE Wirel. Commun. 23(4), 120–127 (2016) 82. B. Li, W. Guo, X. Wang, Y. Deng, Y. Lan, C. Zhao, A. Nallanathan, CSI-independent nonlinear signal detection in molecular communications. IEEE Trans. Signal Process. 68, 97–112 (2019) 83. M.C. Gursoy, M. Nasiri-Kenari, U. Mitra, Towards high data-rate diffusive molecular communications: a review on performance enhancement strategies. Digital Signal Process. 124, 103161 (2021)

References

163

84. H.B. Yilmaz, C.-B. Chae, Simulation study of molecular communication systems with an absorbing receiver: modulation and IsI mitigation techniques. Simul. Modell. Practice Theory 49, 136–150 (2014). [Online]. Available: http://www.sciencedirect.com/science/article/ pii/S1569190X14001452 85. M.K. Zadeh, I.M. Bolhassan, M. Kuscu, Microfluidic pulse shaping methods for molecular communications (2023). Preprint arXiv:2301.05576 86. N. Farsad, N.-R. Kim, A.W. Eckford, C.-B. Chae, Channel and noise models for nonlinear molecular communication systems. IEEE J. Sel. Areas Commun. 32(12), 2392–2401 (2014) 87. M. Kuscu, E. Dinc, B.A. Bilgin, H. Ramezani, O.B. Akan, Transmitter and receiver architectures for molecular communications: a survey on physical design with modulation, coding, and detection techniques. Proc. IEEE 107(7), 1302–1341 (2019) 88. M.O. Araz, A.R. Emirdagi, M.S. Kopuzlu, M. Kuscu, Ratio shift keying modulation for timevarying molecular communication channels (2023). Preprint arXiv:2302.10353 89. M.S. ¸ Kuran, H.B. Yilmaz, I. Demirkol, N. Farsad, A. Goldsmith, A survey on modulation techniques in molecular communication via diffusion. IEEE Commun. Surv. Tutor. 23(1), 7–28 (2020) 90. X. Chen, Y. Huang, L.-L. Yang, M. Wen, Generalized molecular-shift keying (GMoSK): principles and performance analysis. IEEE Trans. Mol. Biol. Multi-Scale Commun. 6(3), 168–183 (2020) 91. D. Jing, Y. Li, A.W. Eckford, Power control for ISI mitigation in mobile molecular communication. IEEE Commun. Lett. 25, 460–464 (2020) 92. R. Jadsadaphongphaibool, D. Bi, Y. Deng, CSK realization for MC via spatially distributed multicellular consortia (2023). Preprint arXiv:2303.12088 93. M.S. ¸ Kuran, H.B. Yilmaz, T. Tugcu, B. Özerman, Energy model for communication via diffusion in nanonetworks. Nano Commun. Netw. 1(2), 86–95 (2010) 94. M.S. Kuran, H.B. Yilmaz, T. Tugcu, I.F. Akyildiz, Modulation techniques for communication via diffusion in nanonetworks, in IEEE International Conference on Communications (ICC), vol. 2011 ( IEEE, Piscataway, 2011), pp. 1–5 95. I. Llatser, A. Cabellos-Aparicio, M. Pierobon, E. Alarcón, Detection techniques for diffusionbased molecular communication. IEEE J. Sel. Areas Commun. 31(12), 726–734 (2013) 96. M.U. Mahfuz, D. Makrakis, H.T. Mouftah, A generalized strength-based signal detection model for concentration-encoded molecular communication, in Proceedings of the 8th International Conference on Body Area Networks (2013), pp. 461–467 97. M.U. Mahfuz, D. Makrakis, H.T. Mouftah, Concentration-encoded subdiffusive molecular communication: theory, channel characteristics, and optimum signal detection. IEEE Trans. Nanobiosci. 15(6), 533–548 (2016) 98. A. Singhal, R.K. Mallik, B. Lall, Performance analysis of amplitude modulation schemes for diffusion-based molecular communication. IEEE Trans. Wirel. Commun. 14(10), 5681–5691 (2015) 99. K.V. Srinivas, A.W. Eckford, R.S. Adve, Molecular communication in fluid media: the additive inverse gaussian noise channel. IEEE Trans. Informat. Theory 58(7), 4678–4692 (2012) 100. A. Einolghozati, M. Sardari, A. Beirami, F. Fekri, Capacity of discrete molecular diffusion channels, in IEEE International Symposium on Information Theory Proceedings, vol. 2011 (IEEE, Piscataway, 2011), pp. 723–727 101. B. Tepekule, A.E. Pusane, H.B. Yilmaz, T. Tugcu, Energy efficient ISI mitigation for communication via diffusion, in 2014 IEEE International Black Sea Conference on Communications and Networking (BlackSeaCom) (IEEE, Piscataway, 2014), pp. 33–37 102. M. Movahednasab, M. Soleimanifar, A. Gohari, M.N. Kenari, U. Mitra, Adaptive molecule transmission rate for diffusion based molecular communication, in 2015 IEEE International Conference on Communications (ICC) (IEEE, Piscataway, 2015), pp. 1066–1071 103. M. Turan, M.S. ¸ Kuran, H.B. Yilmaz, C.-B. Chae, T. Tugcu, Mol-eye: A new metric for the performance evaluation of a molecular signal, in IEEE Wireless Communications and Networking Conference (WCNC), vol. 2018 (IEEE, Piscataway, 2018), pp. 1–6

164

3 Analysis of the Molecular Physical Layer’s Tasks

104. M. Damrath, P.A. Hoeher, Low-complexity adaptive threshold detection for molecular communication. IEEE Trans. Nanobiosci. 15(3), 200–208 (2016) 105. D. Kilinc, O.B. Akan, Receiver design for molecular communication. IEEE J. Sel. Areas Commun. 31(12), 705–714 (2013) 106. B. Li, M. Sun, S. Wang, W. Guo, C. Zhao, Low-complexity noncoherent signal detection for nanoscale molecular communications. IEEE Trans. Nanobiosci. 15(1), 3–10 (2015) 107. V. Jamali, N. Farsad, R. Schober, A. Goldsmith, Non-coherent multiple-symbol detection for diffusive molecular communications, in Proceedings of the 3rd ACM International Conference on Nanoscale Computing and Communication (2016), pp. 1–7 108. A. Gohari, M. Mirmohseni, M. Nasiri-Kenari, Information theory of molecular communication: directions and challenges. IEEE Trans. Mol. Biol. Multi-Scale Commun. 2(2), 120–142 (2016) 109. G. Aminian, H. Arjmandi, A. Gohari, M. Nasiri-Kenari, U. Mitra, Capacity of diffusion-based molecular communication networks over lti-Poisson channels. IEEE Trans. Mol. Biol. MultiScale Commun. 1(2), 188–201 (2015) 110. M.S. ¸ Kuran, H.B. Yilmaz, T. Tugcu, I.F. Akyildiz, Interference effects on modulation techniques in diffusion based nanonetworks. Nano Commun. Netw. 3(1), 65–73 (2012) 111. S. Galmés, B. Atakan, Performance analysis of diffusion-based molecular communications with memory. IEEE Trans. Commun. 64(9), 3786–3793 (2016) 112. N.-R. Kim, C.-B. Chae, Novel modulation techniques using isomers as messenger molecules for molecular communication via diffusion, in 2012 IEEE International Conference on Communications (ICC) (IEEE, Piscataway, 2012), pp. 6146–6150 113. N.-R. Kim, A.W. Eckford, C.-B. Chae, Symbol interval optimization for molecular communication with drift. IEEE Trans. Nanobiosci. 13(3), 223–229 (2014) 114. H. ShahMohammadian, G.G. Messier, S. Magierowski, Optimum receiver for molecule shift keying modulation in diffusion-based molecular communication channels. Nano Commun. Netw. 3(3), 183–195 (2012) 115. N.-R. Kim, C.-B. Chae, Novel modulation techniques using isomers as messenger molecules for nano communication networks via diffusion. IEEE J. Sel. Areas Commun. 31(12), 847– 856 (2013) 116. H. Arjmandi, A. Gohari, M.N. Kenari, F. Bateni, Diffusion-based nanonetworking: a new modulation technique and performance analysis. IEEE Commun. Lett. 17(4), 645–648 (2013) 117. B. Tepekule, A.E. Pusane, M.S. ¸ Kuran, T. Tugcu, A novel pre-equalization method for molecular communication via diffusion in nanonetworks. IEEE Commun. Lett. 19(8), 1311– 1314 (2015) 118. Y. Murin, N. Farsad, M. Chowdhury, A. Goldsmith, Exploiting diversity in one-shot molecular timing channels via order statistics. IEEE Trans. Mol. Biol. Multi-Scale Commun. 4(1), 14– 26 (2018) 119. H. Li, S.M. Moser, D. Guo, Capacity of the memoryless additive inverse gaussian noise channel. IEEE J. Sel. Areas Commun. 32(12), 2315–2329 (2014) 120. A.W. Eckford, Molecular communication: Physically realistic models and achievable information rates (2008). Preprint arXiv:0812.1554 121. C. Rose, I.S. Mian, Inscribed matter communication: part I. IEEE Trans. Mol. Biol. MultiScale Commun. 2(2), 209–227 (2016) 122. Y. Murin, N. Farsad, M. Chowdhury, A. Goldsmith, Time-slotted transmission over molecular timing channels. Nano Commun. Netw. 12, 12–24 (2017) 123. A.W. Eckford, Nanoscale communication with Brownian motion, in 2007 41st Annual Conference on Information Sciences and Systems (IEEE, Piscataway, 2007), pp. 160–165 124. N. Garralda, I. Llatser, A. Cabellos-Aparicio, E. Alarcón, M. Pierobon, Diffusion-based physical channel identification in molecular nanonetworks. Nano Commun. Netw. 2(4), 196– 204 (2011) 125. B.C. Akdeniz, A.E. Pusane, T. Tugcu, Position-based modulation in molecular communications. Nano Commun. Netw. 16, 60–68 (2018)

References

165

126. A. Zare, A. Jamshidi, Receiver design and performance analysis for pulse position modulation technique in diffusion-based molecular communication. Nano Commun. Netw. 21, 100256 (2019) 127. Y. Murin, N. Farsad, M. Chowdhury, A. Goldsmith, Optimal detection for one-shot transmission over diffusion-based molecular timing channels. IEEE Trans. Mol. Biol. Multi-Scale Commun. 4(2), 43–60 (2018) 128. N. Farsad, Y. Murin, A. Eckford, A. Goldsmith, On the capacity of diffusion-based molecular timing channels, in IEEE International Symposium on Information Theory (ISIT), vol. 2016 (IEEE, Piscataway, 2016), pp. 1023–1027 129. N. Farsad, Y. Murin, A.W. Eckford, A. Goldsmith, Capacity limits of diffusion-based molecular timing channels with finite particle lifetime. IEEE Trans. Mol. Biol. Multi-Scale Commun. 4(2), 88–106 (2018) 130. N. Garralda, I. Llatser, A. Cabellos-Aparicio, E. Alarcón, M. Pierobon, Diffusionbased physical channel identification in molecular nanonetworks. Nano Commun. Netw. 2(4), 196–204 (2011). [Online]. Available: http://www.sciencedirect.com/science/ article/pii/S1878778911000378 131. Y.-P. Hsieh, Y.-C. Lee, P.-J. Shih, P.-C. Yeh, K.-C. Chen, On the asynchronous information embedding for event-driven systems in molecular communications. Nano Commun. Netw. 4(1), 2–13 (2013). [Online]. Available: http://www.sciencedirect.com/science/article/ pii/S1878778912000543 132. B. Krishnaswamy, C.M. Austin, J.P. Bardill, D. Russakow, G.L. Holst, B.K. Hammer, C.R. Forest, R. Sivakumar, Time-elapse communication: bacterial communication on a microfluidic chip. IEEE Trans. Commun. 61(12), 5139–5151 (2013) 133. N. Farsad, Y. Murin, W. Guo, C. Chae, A.W. Eckford, A. Goldsmith, Communication system design and analysis for asynchronous molecular timing channels. IEEE Trans. Mol. Biol. Multi-Scale Commun. 3(4), 239–253 (2017) 134. J.G. Gonzalez, J.L. Paredes, G.R. Arce, Zero-order statistics: a mathematical framework for the processing and characterization of very impulsive signals. IEEE Trans. Signal Process. 54(10), 3839–3851 (2006) 135. M.C. Gursoy, D. Seo, U. Mitra, Concentration and position-based hybrid modulation scheme for molecular communications, in ICC 2020-2020 IEEE International Conference on Communications (ICC) (IEEE, Piscataway, 2020), pp. 1–6 136. N.V. Sabu, N. Varshney, A.K. Gupta, On hybrid MOSK-CSK modulation based molecular communication: Error rate performance analysis using stochastic geometry (2019). Preprint arXiv:1904.09736 137. M.C. Gursoy, D. Seo, U. Mitra, A concentration-time hybrid modulation scheme for molecular communications. IEEE Trans. Mol. Biol. Multi-Scale Commun. 7(4), 288–299 (2021) 138. Y. Huang, M. Wen, L.-L. Yang, C.-B. Chae, F. Ji, Spatial modulation for molecular communication. IEEE Trans. Nanobiosci. 18(3), 381–395 (2019) 139. G.M. Karthiik, S. Joshi, A novel hybrid modulation scheme for molecular communication: Performance analysis. IEEE Wirel. Commun. Lett. 11, 1234–1238 (2022) 140. L. Brand, M. Garkisch, S. Lotter, M. Schäfer, A. Burkovski, H. Sticht, K. Castiglione, R. Schober, Media modulation based molecular communication (2022). Preprint arXiv:2203.00722 141. M.S. Kopuzlu, M.O. Araz, A.R. Emirdagi, M. Kuscu, Capacity analysis of molecular communications with ratio shift keying modulation (2022). Preprint arXiv:2205.13317 142. M. Ahuja, M.R. Bhatnagar, A spatio-temporal coded modulation scheme for molecular mimo systems, in 2021 International Conference on COMmunication Systems & NETworkS (COMSNETS) (IEEE, Piscataway, 2021), pp. 140–144 143. M. Wen, F. Liang, Y. Tang, Layered molecular shift keying for molecular communication via diffusion. IEEE Commun. Lett. 25(11), 3599–3603 (2021) 144. X. Wang, Z. Jia, A new modulation method for diffusion molecular communication. Acad. J. Eng. Technol. Sci. 4(2), 7–12 (2021)

166

3 Analysis of the Molecular Physical Layer’s Tasks

145. O. Kara, G. Yaylali, A.E. Pusane, T. Tugcu, Machine learning based molecular index modulation (2021). Preprint arXiv:2103.09812 146. M. Ahuja, M.R. Bhatnagar, A pulse position-based spatio-temporal coded modulation for molecular mimo systems. IEEE Commun. Lett. 26, 2370–2374 (2022) 147. M. Egan, B.C. Akdeniz, B.Q. Tang, Stochastic reaction and diffusion systems in molecular communications: recent results and open problems. Digital Signal Process. 124, 103117 (2022) 148. M.N. Anjum, H. Wang, Molecular communication for wireless body area networks (2019) 149. A. Mukherjee, S. Das, S. Chatterjee, An architecture of calcium signaling for molecular communication based nano network, in Modeling, Methodologies and Tools for Molecular and Nano-scale Communications (Springer, Berlin, 2017), pp. 165–203 150. Y. Cevallos, L. Molina, A. Santillán, F. De Rango, A. Rushdi, J.B. Alonso, A digital communication analysis of gene expression of proteins in biological systems: a layered network model view. Cognit. Comput. 9(1), 43–67 (2017) 151. Y. Cevallos, L. Tello-Oquendo, D. Inca, C. Palacios, L. Rentería, Genetic expression in biological systems: a digital communication perspective. Open Bioinformat. J. 12(1), 45–49 (2019) 152. Y. Cevallos, T. Nakano, L. Tello-Oquendo, D. Inca, I. Santillán, A.Z. Shirazi, A. Rushdi, N. Samaniego, Modeling gene expression and protein delivery as an end-to-end digital communication system. Open Bioinformat. J. 14(1), 21–35 (2021) 153. Y. Cevallos, T. Nakano, L. Tello-Oquendo, N. Chopra, A.Z. Shirazi, D. Inca, I. Santillán, Theoretical basis for gene expression modeling based on the ieee 1906.1 standard, in International Conference on Bio-inspired Information and Communication Technologies (Springer, Berlin, 2021), pp. 145–162 154. J. Wang, M. Peng, Y. Liu, X. Liu, M. Daneshmand, Performance analysis of signal detection for amplify-and-forward relay in diffusion-based molecular communication systems. IEEE Int. Things J. 7(2), 1401–1412 (2019)

Chapter 4

Case Studies of Applications of Digital Networks Theories to Molecular Network Stacks

The development of technology in recent years has made it possible for various scientific disciplines to collaborate and create effective models that utilize a mimic of the true nature and its traits. Examples of these fields include engineering and biology. These models can be used in a variety of fields, including business, engineering, medical, biochemistry, biotechnology, computer sciences, and other fields. For instance, bionanotechnology is a subfield of nanotechnology that incorporates biological materials, makes use of biological design or manufacture, and is used in biotechnology or medicine [1–4]. The molecular communication (MC) branch of bionanotechnology, on the other hand, has drawn a lot of interest. It is a relatively new method that employs biochemical signals to achieve information exchange among artificially and naturally made bio-nanoscale machines across short distances. The need for genomic signal processing is also rapidly increasing due to its vital role in the treatment of diseases and the significance of human genetics and related fields [5, 6]. For instance, several MC apps concentrate on the identification and management of illnesses. In reality, individualized forecasts of disease evolution are possible through the emulation of nanoscale biological processes. This is accomplished, for instance, by identifying a small number of biomarkers, which limits the use of conventional in vivo testing on patients and aids medical professionals in determining the best treatments. Based on a thorough assessment of the MC characteristics utilizing the findings from in vitro and in vivo research, the processing models were developed. Also, a wide range of disorders could be detected and treated, thanks to the development of nanosensors and actuators (e.g., cardiovascular diseases, tumors, among others). These nanodevices could initiate drug delivery systems and/or immune system activation in limited, focused locations without harming the rest of the body (i.e., smart drugs) [7–10]. It is also important to note that MC systems can be used in a variety of different industries, including food production, the development of functional materials and clothing, as well as military and environmental applications [11]. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 Y. Cevallos et al., Molecular Communications, https://doi.org/10.1007/978-3-031-36882-0_4

167

168

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

In light of what was said in the preceding sentence and the networking theories used to study MCs in earlier chapters, we will now investigate some significant examples of these related notions.

4.1 Case 1. Bacterial Molecular Communication Based Nanonetworks NanoNS3, a network simulator built on Network Simulator 3, is described in [12, 13] for modeling bacterial molecular communication (BMC) networks. In order to simulate BMC networks, nanoNS3 implements some of the fundamental protocols. The receiver response model, the channel loss model, the on-off keying model, and the amplitude addressing model are the four crucial models that nanoNS3 has developed. Figure 4.1 depicts the high-level structure of nanoNS3. In the same illustration, the names of seven significant classes are listed together with

Fig. 4.1 NanoNS3 architecture

4.2 Case 2. Internet-of-Nano-Things Healthcare Applications

169

their respective network-layer structures. Below, a brief discussion of each class’s functionality is provided [12, 13]: • NanoNetDevice. It is comparable to the network interface card (NIC), and it can support various nano-communication technologies and associated protocols, such as diffusive or electromagnetic wave-based nano-communication schemes (e.g., amplitude addressing). • NanoNode. Different NanoNetDevices can be integrated with NanoNode to provide relevant communication technologies and protocols, allowing NanoNode to communicate with each other. It can be thought of as the physical device [12, 13]. • PacketSocket. This is a straightforward and unique NS-3 application class that does not make use of IP (Internet Protocol) addresses. By adjusting applicationrelated parameters, such as packet arriving interval, the number of maximal transmission packets, and packet size, it is utilized to set up user-defined applications for nano-communications [12, 13]. • NanoMessage. By adjusting application-related parameters, such as a packet arriving interval and packet size, this class is used to configure user-defined applications for nano-communications [12, 13]. • NanoRouting. Each NanoNode’s message forwarding is controlled by this class [12, 13]. • NanoMAC. This class controls the channel access of various NanoNodes and the MAC layer addressing scheme [12, 13]. • NanoPHY. This class is used to simulate the transmission and reception of nano signals by transmitters and receivers. Modulation, demodulation, and receiver response are among the features this class is responsible for [12, 13]. • NanoChannel. In order to mimic how transmitted signals are propagated and attenuated in the associated microfluidic channel, this class is used to build up the channel settings. The channel loss may then be determined [12, 13].

4.2 Case 2. Internet-of-Nano-Things Healthcare Applications The biological characteristics of the tissues or organs that nanonetworks are monitoring can be regularly sensed, and readings can be sent to gateways. Nanonetworks are also capable of sending alarms when certain compounds, chemicals, or viruses are detected. Moreover, detection can be more advanced and concentrate on anomalies, as shown in Fig. 4.2 [14]. A generic IoBNT (Internet-of Bio-Nano-Things) architecture for universal healthcare aids in the monitoring of tissues or organs, such as the signs of a heart attack beginning. Nanoscale or intracellular regenerative tissue engineering, drug administration, and other proactive remedial operations inside the body can all be carried out via nanonetworks. The IoBNT has a layered architecture with the following primary functionalities in each layer [14, 15]:

170

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

Fig. 4.2 Generic IoBNT architecture for ubiquitous healthcare

(a) Application Layer. Applications for regulating and monitoring medical occurrences in real-time or almost real-time are adequately run by the application layer [14, 15]. (b) Transport Layer. Because the body is thought of as a noisy environment for the transmission of information, the transport layer enhances dependability through the widespread usage of nanodevices with the same functionality and the ability to report the same data [14, 15]. (c) Network Layer. Because of the drift of the communication medium, which creates latency, and the random positioning of the nanodevices, the network layer must offer routing and multi-hops in the network, which is challenging. As a result, proximity-based opportunistic routing and device mobility inside the body may be appropriate in these networks [14, 15]. (d) Medium Access Control/Physical Layer. Health applications need three key components from this layer: a channel capacity that ensures reliable data delivery, an accurate channel model that takes into consideration the particular biological transmission medium and its corresponding noise, and effective coding schemes that are error-resistant [14, 15].

4.3 Case 3. Modeling Nonviral Gene Delivery as a Macro-to-Nano. . .

171

4.3 Case 3. Modeling Nonviral Gene Delivery as a Macro-to-Nano-Communication System Gene delivery systems that transmit genetic information encoded as DNA (deoxyribonucleic acid) to living cells can be considered communication systems because the primary function of any communication system is to deliver information from a source to a sink. As a result, the authors in [16] establish that modeling methods created for conventional communication systems should be transferable to gene delivery systems. Thus, [16] proposes a nonviral gene delivery method in which the interactions between the macro-scale (an application) and nano-scale begin as soon as this application decides that a cell has to undergo genetic modification or manufacture a specific protein through: (a) Encoding of information in a format understood by the destination and protecting its integrity during the transport [16]. (b) Delivery of the information carriers through the cell membrane—a process called internalization—into endosomes [16]. (c) Routing the information carriers through cytoplasm [16]. (d) Delivery of the carriers to cell nucleus [16]. (e) Reception and decoding of information by the nucleus [16]. (f) Invoking an expected application resulting in a modified cell behavior or production of the requested protein [16]. Hence, a protein of interest that encodes that information is created using plasmid DNA, and the cell nucleus will decode it. When cationic polymers (polyplexes) or lipids (lipoplexes) are used to create nanoparticles known as complexes, they are creating message carriers or transport PDUs (protocol data units). These plasmids encode exogenous genes that are to be introduced into cells. DNA is shielded against deterioration by nucleases and serum constituents through complexation. By lowering the effective size of DNA and encouraging interactions between positively charged DNA complexes and the negatively charged cellular membrane, it improves cellular absorption. Complexes are the PDUs for the network layer, and the complexation process can be thought of as network layer framing. DNA complexes are generated and then transported to cells in solution. The solution with complexes is delivered to the media surrounding the cells via the physical layer operation at the macroscale. The initial node of the communication chain involved in nonviral gene distribution is the layered protocol stack for the macro host, which is depicted in Fig. 4.3 and described in [16]. Complex binding to the cell can happen through receptor binding or nonspecific binding after the addition of complexes to the media. Hence, the system’s first node at the nanoscale can be thought of as an endosome (Fig. 4.3). In order for a complex

172

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

Fig. 4.3 Gene delivery process described using a four-layer communication protocol

to be internalized into the cell, an endosome must first be formed around it. From there, the complex can either be sent to a lysosome for breakdown or released into the cytoplasm. This is how the physical layer of the endosome operates. Endosomes thus do a type of routing. Endosome function can be viewed as implementing two communication protocol layers, the physical layer and the network layer, as they carry out processing up to complicated routing. Complexes are encapsulated in endosomes by the physical layer of endosomes, which is also in charge of releasing complexes into the cytoplasm or delivering them to lysosomes for destruction. The network layer of endosomes makes the decision regarding the location of the encapsulated complexes [16]. The cytoplasm serves as an intermediary node on the pathway to the nucleus, directing incoming messages through routing complexes to the nucleus or directing plasmids after they have been unpacked from complexes (lipids or polymers). Complexes that are being routed bind to cytoplasmic proteins that carry nuclear localization sequences in the cytoplasm, and these proteins then transport the complex into the nucleus. This process is comparable to adding a header with the destination address to a packet, which occurs at the network layer of the communication protocol [16]. In the second approach (Fig. 4.3), the complexes are first unpacked (plasmid dissociation from cationic polymer or lipid), and the same binding to nuclear localization proteins (addition of network layer headers) operation takes place. The packaged plasmids, which are the transport layer PDUs, are however degraded prior to binding. This degradation can be viewed as a type of integrity check carried out at the transport layer of the cytoplasm for modeling purposes [16]. Nuclear pores, which serve as an interface between the nucleus and cytoplasm and are therefore regarded as the physical layer of the nucleus, allow both plasmids and complexes to enter the cell nucleus. During mitosis, the nuclear barrier breaks down, allowing them to enter the nucleus through random diffusion. Then, as processing at the Network Layer of the nucleus, complexes and plasmids are released from nuclear localization proteins in an action that is the opposite of the

4.3 Case 3. Modeling Nonviral Gene Delivery as a Macro-to-Nano. . .

173

binding process that takes place in the cytoplasm (network layer of cytoplasm). Plasmids are prone to interference and noise (i.e., biological deterioration) on their journey from macro-scale preparation to effective delivery to the cell nucleus, which causes transmission mistakes. Only plasmids that pass some sort of integrity test in the nucleus are able to transfect cells. These functioning plasmids are known as error-free plasmids. In fact, one could say that a virtual transport channel exists between the macro-scale and the cell nucleus and that the integrity checking is a transport layer function of the cell nucleus [16]. The nucleus of the cell begins manufacturing the appropriate mRNA as soon as the application layer of the nucleus recognizes an active plasmid (messenger ribonucleic acid). Successful detection of an active plasmid by the nucleus’ application layer triggers transcription, which decodes the plasmid’s DNA and generates the necessary RNA (ribonucleic acid) molecules. As long as there are active plasmids inside an intact cell nucleus, the RNA molecules, here referred to as PDUs of the nuclear application layer, are then transformed into mRNA and transported into the cytoplasm [16]. The ribosomes, which are primarily found on the endoplasmic reticulum, translate the mRNA molecules into the corresponding chain of amino acids, which results in an unfolded protein after they have been created in the nucleus and transported into the cytoplasm. A portion of the mRNA gets destroyed prior to translation as a result of noise and interference that affect the mRNA molecules. This integrity check and translation can be thought of as a transport layer protocol operation [16]. After being folded into a mature protein, either the produced amino acid chain is kept in the cell (nucleus, cytoplasm, other organelles) for later use by the corresponding application (participation in metabolism, cell signaling, etc.) or the protein may be secreted from the cell to perform extracellular functions (Fig. 4.3). As an alternative, it might already be damaged. As the destination address is appended to the data packet during folding and/or binding to chaperones, this procedure can be viewed as a network layer protocol operation. The developed proteins are subsequently used by the relevant applications or, in the worst-case scenario, degraded. The rate of protein degradation for mature proteins also accounts for the breakdown of misfolded proteins that are irreparably damaged. When network PDU headers contain unrecoverable faults, packets are dropped. This degrading process is similar to that [16]. Biological communication systems are evaluated in [17] using an example from layered communication protocols in telecoms to show the transmission of HIV (human immunodeficiency virus). Thus, a variety of techniques are applied to HIV in the research in [17]. Because the HIV virus wants to spread as widely as possible, it replicates itself in the communication channel. This fact is quite similar to a multicast communication network, where there are numerous receivers (called cells) that can receive packets (viruses). Figure 4.4 (as a paradigm of a macro- to microlevel communications system) shows the HIV infection-layered protocols, whose effectiveness is given in [17].

174

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

Fig. 4.4 HIV infection-layered protocol

In this HIV infection and multicast communication network, each individual virus is essentially delivered to a data channel, with each target cell acting as the receiver that can take in this packet. There is a chance that a certain cell will become infected and start producing new viruses at a consistent rate (called effective infected cells). At this point, the cells alter their behavior patterns and begin to copy and send the packet to the channel in addition to receiving it. According to the protocol stack, a new virus spreads to other cells by way of a communication network. For this to happen, the packet must first transit via the channel before it can be effectively received. When a cell is infected, information is transmitted (or a protocol is started); nevertheless, even though the transmission of information (or a protocol) starts at this node, the virus still needs to penetrate the system. To put it more simply, the complete communication process must begin with a single initial data packet transmitted and entering the channel. After virus creation, the virus must leave host cells to engage in routing or pass through a network layer inside the cell [17]. In order to reach the target cells, the virus then travels along the channel. The virus creates a new stage of the route when it reaches the cell’s surface, and this stage is again represented by a network layer for communication. The virus must be absorbed once it enters the target cell. The virus can travel by fusion, endocytosis, and breakdown, among other processes. When a virus infects a cell by fusing, its components are injected after adhering to the surface of the target cell. In the case of endocytosis, the virus is consumed by the cell. Phagocytic cells that use the viralencapsulating vesicles to eat the virus are involved in this process. The virus will probably die if it is unable to get rid of the vesicles. On the other hand, the virus can infect cells if it escapes from the vesicle. The virus must attempt again through

4.4 Case 4. Internet of Things for Advanced Targeted Nanomedical Applications

175

fusion or endocytosis into cells if it dissociated from cells. There might be a virus there that isn’t properly connected to the cells and can’t send messages. In addition to routing during entry, additional routing and error-control procedure is carried out after the virus enters the cell via endocytosis [17]. Once the virus’s components have been effectively ingested, it must assemble its contents to generate its message (RNA). Reverse transcription, or the process of turning RNA into DNA, is incredibly error-prone; it is quite possible that the DNA may mutate or assemble wrongly to send the wrong message. It is obvious that error control is used to assist the virus in putting together its message so that it can be effectively received. The protocol finally explains how DNA travels to the cell nucleus. Although there are many variables that can affect this strategy’s effectiveness, it is actually another type of routing. The virus starts replicating in the cell as soon as it penetrates the nucleus and the message is activated. A clear indication must be interpreted by cells and implemented, which requires that the message be accurately provided to the application layer. The communication channel is deemed finished at this point. The process starts over at the application layer in the host cells as soon as the cells begin to manufacture a new virus. This communication method basically forms a loop between the two application levels of the protocol and is a continuous cycle [17].

4.4 Case 4. Internet of Things for Advanced Targeted Nanomedical Applications The main goal of nanomedicine is to use nanotechnology concepts, tools, technologies, and methods to increase the effectiveness of medical and healthcare systems. On the other hand, MC engineering, a different branch of nanotechnology, focuses on creating nanoscale machines and devices that can share biological information [18, 19]. The potential for the medical and healthcare systems is increased by incorporating the idea of MC into nanomedicine, which enables the coordination of operations and information sharing among numerous nanomedical machines and devices [20– 22]. With regard to nanotechnology and healthcare delivery, the concept of the IoT (Internet of Things) has ushered in related concepts such as the Internet of NanoThings (IoNT) [12, 23–26], the Internet of Bio-NanoThings (IoBNT), and Internet of Bio-NanoThings for Ambient Assisted Living (IoBNTAAL) into the research and industrial domain. The IoNT takes into account the potential for creating nanomachines that can communicate via the Internet. The IoBNT envisions a future application domain in which the actions of nanosystems working within an in-body nanonetwork can be tracked and managed online [27]. The IoBNTAAL specializes in the IoBNT to ambient assisted living [18]. In order to deliver individualized nanomedical therapy, monitoring, and control anywhere, anytime, and for anybody, the IoT-ATN (Advanced Targeted

176

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

Fig. 4.5 IoT-ATN-layered architecture

Nanomedical Applications) leverages the advantages of the IoNT, IoBNT, and IoBNTAAL. To query the states of a collection of living and nonliving systems in the IoT-ATN system, and to change those states, there is a need for embedded sensors, actuators, processors, and transceivers, all of which must cooperate. To accomplish this, the IoT-ATN solution must ensure effective connectivity among the IoT-ATN devices and services through a layered architecture, as seen in Fig. 4.5 and analyzed as follows [18]: • Environment of Things Layer. This layer consists of the things (patient or patients whose bodies the nanosystems and sensors/other devices’ function) or locations (the patient’s immediate environment) where the nanosystems and sensors’ observations and activities take place [18]. • Device Layer. Hardware like nanosystems, body-area sensors, relays, environmental sensors, tags, readers, and embedded systems are taken into account at this physical layer. In the IoT-ATN, body-area sensors/actuators, along with other devices, establish an in-body nanonetwork with nanosystems. The nanosensors in the nanonetwork gather data on the various conditions of the patient’s internal

4.5 Case 5. Hybrid DNA- and Enzyme-Based Computing for Address. . .

•

•

•

•

177

environment, cells, tissues, and organs, as well as the operations of other nanosystems in the nanonetwork, inside the body. The body-area network of the IoT-ATN is made up of various devices, such as body-area sensors and actuators, which collect patient biosignal data from the outside. Environmental sensors that are built into the off-body network can also be used to gather sensory data on things like temperature, humidity, air quality, and movement. For the storing of information, tags may be situated in the off-body network or in the body-area network. Moreover, centralized or distributed embedded edge processors will typically be used to process acquired data [18]. Connectivity and Data Processing Layer. The different nanosystems in the nanonetwork, the devices in the body-area network, and the devices in the offbody networks are all handled by this layer, which also handles communication connectivity and data processing. The physical positioning and configuration of the network equipment to meet the desired objectives are issues at this layer. Also, this layer is responsible for processing the data that is transferred between the various networks in a way that makes it possible to gather and process information in real time as needed [18]. Gateway and Network Layer. This layer serves as the fundamental link between the Internet, body-area network, off-body network, and nanonetwork. This layer takes into account issues including routing and addressing, network and transport capabilities, and error detection and repair. Certain varieties of nanogateways, microgateways, and macrogateways will be in charge of these tasks. This layer ought to be capable of managing connectivity in a diverse network [18]. Management and Service Layer. Within the IoT-ATN environment, this layer supervises and coordinates the many network service providers. Its duties include a range of services and include data and application administration, control, security, monitoring, storage, organizing, and visualization. Different policies, including quality of service management, traffic management, device management, traffic engineering, business process modeling/execution, packet inspection, identity management, access authorization, and functions, are taken into consideration in this layer to help manage the information generated by all of the devices and network constituents [18]. Application Layer. In the IoT-ATN context, this layer is in charge of supplying a variety of apps that offer user interfaces. It outlines the use cases for the IoT-ATN, in this case, the provision of individualized medical care [18].

4.5 Case 5. Hybrid DNA- and Enzyme-Based Computing for Address Encoding, Link Switching, and Error Correction in Molecular Communication Using DNA and enzyme computing, the authors of [28] and [29] present a novel method for creating MC protocols. As shown in Fig. 4.6, they create a straightforward protocol stack for MC in which each protocol contributes to a particular task

178

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

Fig. 4.6 DNA- and enzyme-based protocol

to support addressing, error correction, and link switching. Vesicles and calcium signaling are used as an MC mechanism to transfer the message molecules at the physical layer. While enzymatic computation is employed to carry out a small-size logic circuit with high-speed calculation, the physical layer functions also include methods like modulation and channel coding. More complex computations like FEC (forward error correction), addressing, and encoding/decoding will be needed at the transport and application layers. It is possible to use the inter-layer protocol between them to control the computation process and position within the cell.

4.6 Case 6. Efficient Molecular Communication Protocol Based on Mobile Ad Hoc Nanonetwork A hybrid communication paradigm for long-distance transmission in [28] allows the construction of an MC protocol (MCP) in the molecular network layer. Figure 4.7 depicts the three nanomachines that make up the suggested communicational scenario: a transmitter, a destination, and a relay. The nanomachine-based carrier S molecules that the transmitter sends travel through the surroundings until they arrive at their destination. The relay, which is situated between the transmitter and the destination, functions as a control node and can detect molecules being transferred as well as feedback or loss resulting from the destination’s chemical response. Enzymatic kinetic states that the MCPbased feedback scheme, whether positive or negative, has the ability to modify

4.6 Case 6. Efficient Molecular Communication Protocol Based on Mobile Ad. . .

179

Fig. 4.7 Molecular layered architecture for illustrating the flow of molecules information from the T x to Dx by using relay .(Rx)

the transmission rate in order to obtain the highest throughput and efficiency. The goal throughput and efficiency between the two transmitters and the destination can be maintained via feedback mechanisms. When the distance from the sender grows, the concentration of S-type molecules drops. In order to enhance information throughput, the sender broadcasts at a high transmission rate. But if the receiver does not react quickly enough to process the molecules that are arriving, some molecules will be lost in the channel, which reduces the efficiency of the nanonetwork [28]. In order to get around this issue [28], a nanomachine serving as a relay is reportedly able to detect the concentration of molecules being transferred at the transmitter and the molecules being lost at the destination. The expression for the reaction rate at the destination is given by .

Iout (R) = v.J ([S](R))

(4.1)

where .Iout is the receiver’s reaction rate, v is the volume of the receptor on the receiver and .J ([S](R)) is the reaction rate at the receiver of the concentration of S molecules [28]. The advantage of the suggested MCP is using a straightforward protocol as opposed to a complicated receiver or transmitter. The issues with message handling are resolved by transferring the message straight from the physical layer to the link layer. It is challenging to group a biological machine at the link layer if it releases individual signal molecules at the physical layer. Also, it can be challenging to identify a bio-nanomachine at the relay and the separate header from it if it releases a group of signal molecules. Furthermore useful for regulating chemical packet propagation between source and destination is routing. It permits long-distance packet transmission and the ability to send packets with various messages to various destinations. Moreover, it permits the link layer to store the molecular frames. The primary goal of the routing process is to concentrate on problems in a single layer. It also enhances a few system parameters, such as throughput and efficiency, and enables global optimization (Eqs. 4.2 and 4.2). The relay uses a straightforward

180

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

computational method, like a molecular NAND gate, to transport information at a low or rapid pace. The sender receives the feedback letter F. According to the proposed classification, there are two sorts of this feedback: NF feedback and PF feedback. As the concentration of the feedback molecules rises at NF, the sender bio-NM slows down transmission. As the number of feedback molecules grows at PF, the sender speeds up the transmission. Based on the concentration of the feedback at the sender location, the sender modifies the transmission rate (Eq. 4.4) [28]: .

μmax = vMJ (Css (r)); Css (r) =



N Iin



2 kD exp(

.

ηmax =

,k > 0 k D r)

vMJ (Css (r)) N Iin

(4.2) (4.3)

where .J (.) is the chemical reaction rate that is computed on the basis of kinetic properties, N is the number of the sender nanomachines, v is the volume of the receptor on the receiver, M is the number of the receiver nanomachines, .Iin is the rate of S molecules transmitted by sender, .CSS is the steady-state concentration .(at t → ∞) of S type molecule at distance r, k is degradation rate and finally D is diffusion coefficient [28]: .

Ini (t) = vs js ([F ](0, t))

(4.4)

where .vs is the sender volume, .js (.) is the sender reaction rate, .[F ](0, t) is the feedback molecules at the sender location [28].

4.7 Case 7. IEEE Standard Data Model for Nanoscale Communication Systems In its first version, the IEEE (Institute of Electrical and Electronics Engineers) 1906.1 standard recommended practice (approved in 2015 [30]) contains a conceptual model and standard terminology for ad hoc network communication at the nanoscale. More specifically, this recommended practice contains the following: (a) the definition of nanoscale communication networking; (b) the conceptual model for ad hoc nanoscale communication networking; and (c) the common terminology for nanoscale communication networking, including (1) the definition of a nanoscale communication channel highlighting the fundamental differences from a macroscale channel; (2) abstract nanoscale communication channel interfaces with nanoscale systems; (3) performance metrics common to ad hoc nanoscale communication networks; and (4) the mapping between nanoscale and traditional communication networks, including necessary high-level components such as a map of significant components: coding and packets, addressing, routing, localization, layering, and

4.7 Case 7. IEEE Standard Data Model for Nanoscale Communication Systems

181

Fig. 4.8 Example OSI to nanoscale communication network mapping

reliability. In [31], a new version of the standard was developed, i.e., 1906.1.1 (approved in 2020 by Internet Engineering Task Force, Request for Comment 7950) in which IEEE establishes a common YANG (Yet Another Next Generation) data model for IEEE 1906.1 nanoscale communication systems. The IEEE 1906 framework can be compared or mapped into communication protocols. There have been many attempts to map nanoscale communication networks to the Open Systems Interconnection (OSI) model (ISO/IEC standard 7498-1:1994) in different ways, as an informative example (Fig. 4.8). Due to their size, nanoscale communication systems are simpler and less easily programmed than macroscale systems. The IEEE 1906 framework is representative of those found in natural, small-scale settings, such as biological systems. The result is less control with which to implement the details of OSI logical layers. For example, perturbation is simpler and has a tendency to be more mechanical in implementation. Motion is more random, for example, subject to Brownian forces. Some type of gradient or field is required to improve motion. There is also less control over specificity (addressing) [30, 32]. OSI model (ISO/IEC 7498-1 [33]) defines the traditional seven-layer protocol stack. The IEEE 1906 framework can be considered to reside within the lowest layers of the OSI stack. More specifically: (a) Message relates approximately to a classical frame, packet, or protocol data unit (PDU) [30–32]. (b) Message Carrier (Component 0) relates to a wave (the characteristic of a wave that encodes information) [30–32]. (c) Motion (Component 1) relates approximately to the classical physical layer (wave propagation) [30–32].

182

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

(d) Field (Component 2) relates approximately to the classical data link and network layers (ensuring node-to-node information flow) [30–32]. (e) Perturbation (Component 3) relates approximately to classical modulation at the physical layer [30–32]. (f) Specificity (Component 4) relates approximately to classical addressing at the data link layer [30–32]. (a) Component 0: Message Carrier The message carrier provides the service of transporting the message. The message carrier may be either particle or wave. Similar to quantum mechanics, the message carrier may also be a simultaneous combination of both particle and wave. Molecular structure may encode information transported by the message carrier from a transmitter to a receiver. Wavelike changes in message concentration may also encode information. In 1906.1.1 (Section 5.3.3.3.2) are defined some core message carrier specifications (with a label name) which include molecular motors [30–32]. (b) Component 1: Motion Defines the movement capability for Component 0 (Message Carrier). The motion component provides the service of movement for the message carrier (in any direction) caused by force or thrust applied to the message carrier. Motion provides the necessary potential to transport information through a communication channel. Message carriers can be active, generating their own motion, or passive, being propagated by the media. Examples include molecules diffusing through fluids, Brownian motion, and self-propelled motion. Motion (Component 1) relates approximately to the classical physical layer (wave propagation). In 1906.1.1 (Section 5.3.3.3.2) are defined some motion specifications (with a label name) which include movement through diffusion in this component [30–32, 34]. (c) Component 2: Field Defines organized movement of Component 1 (Motion). The field component provides the service of the organized motion for message carriers. It can be thought of as a virtual waveguide in communications. The field may be implemented internally or externally relative to the medium. Examples include an internal implementation including swarm motion or flocking behavior; external implementations are nonturbulent fluid flow, EM field, chemical gradient released to guide movement of bacteria, and molecular motors guided by microtubules. Field (Component 2) relates approximately to the classical data link and network layers (ensuring node-to-node information flow). In 1906.1.1 (Section 5.3.3.3.2) are defined some field specifications (with a label name) which include concentration gradient in this component [30–32]. (d) Component 3: Perturbation Defines the signal transported by Component 0 (Message Carrier). The perturbation component provides the service of varying message carriers as needed to represent a signal. This may be thought of as modulation (signal impression).

4.7 Case 7. IEEE Standard Data Model for Nanoscale Communication Systems Table 4.1 IEEE 1906.1 core component specification for 1906.1.1 standard for nanoscale communication systems

Label (name) Bacterium Calcium-ion Charge Electromagnetic-wave-variation Ligand Motor Diffusion Motion-type Potential-difference Walking Wave-guide Compartmentalized Concentration-gradient Directional-antenna Microtubules Nanostructure-orientation Concentration-change Electrical-current-variation Molecular-structure Transmission-rate Antenna-aperture Receptor-sensitivity Electrical-charge

183

Parent component Message-carrier

Motion

Field

Perturbation

Specificity

Examples include signals based on the number of received message carriers, controlled dense-versus-sparse concentrations of molecules, the simple onversus-off flow of signal molecules, using different types of message carriers, and modifying the conformation of molecules (e.g., DNA) to represent multiple states. In 1906.1.1 (Section 5.3.3.3.2) are defined some perturbation specifications (with a label name) which include molecular structure in this component. Perturbation (Component 3) relates approximately to classical modulation at the physical layer [30–32]. (e) Component 4: Specificity Defines targeted reception of Component 3 (Perturbation). The specificity component provides the service of sensing or reception of a message carrier by a target. This can be mapped to addressing in classical communication systems. Examples include the shape or affinity of a molecule to a particular target and complementary DNA for hybridization. In 1906.1.1 (Section 5.3.3.3.2) are defined some specificity specifications (with a label name) which include receptor sensitivity in this component. Specificity (Component 4) relates approximately to classical addressing at the data link layer [30–32] (Table 4.1).

184

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

4.8 Case 8. Gene Expression and Protein Delivery Analysis from a Network-Layered Paradigm The architecture for MC may be discussed from a computer network perspective the TCP/IP (Transmission Control Protocol/Internet Protocol) reference model. As an example, Fig. 2.1 illustrates how information may flow from a source through a router to the destination, representing groups of sender bio-nanomachines, bionanomachines with routing functionality, and receiver bio-nanomachines, respectively. The source and router, as well as the router and destination, are within a communication range, meaning that information molecules can propagate from one to the other within a reasonable amount of time to induce the intended reactions. Similar to the TCP/IP reference model, the application layer provides a set of options to implement applications; the network and link layers provide mechanisms to transmit information over and within a communication range; and the physical layer provides biophysical mechanisms for transmission, propagation, and reception of information molecules over physical media. The network layer at the source selects a communication channel, and the link layer ensures that the channel is available. The physical layer transmits and propagates information molecules over the selected channel to the router. The router then similarly selects a communication channel, ensures that the channel is available, and transmits and propagates a type of information molecule to the destination [32]. With this ICT (information and communication technology) bases in [11, 32, 35– 43] is applied network theories (e.g., independent functions of layers in a stack, addressing, flow control, error control, and traffic control) in the analysis of gene expression (transcription and translation of DNA). (a) Transcription of DNA The basic model of MC may be described based on Shannon’s model of communication (i.e., transmitter, communication channel, and receiver) and as a nanoscale communication network is a human-designed system for communicating at or with the nanoscale, using physical principles that are suited to nanoscale systems; then, from the works proposed in [32, 35–37], the authors suppose that cellular nucleus represents the biological transmitter (or a biological DTE (Data Terminal Equipment) that contain the information source) composed of nucleotide blocks called genes (which must be processed intracellularly or extracellularly). In [32, 35–37] authors particularly focus into a long-distance cellular communication (exhibited, e.g., in the endocrine system, this type of MC is known as long nanorange communication). In nature a gene is a set of nucleotides that stores the information required for accomplishing a specific function (by a protein or RNA) to be performed at a specific destination; thus, they hypothesize that the contents of a gene can be understood as a type of addressing at the network layer. The transmission of information from the nucleus cell (i.e., from genes) to a certain destination begins with the transcription process in which the information of DNA is copied into RNA [32, 35–37].

4.8 Case 8. Gene Expression and Protein Delivery Analysis from a Network. . .

185

The DNA molecules contain digital information as it is encoded by four discrete values (four nucleotides, of information carried by a  the quantity  single nucleotide base is 2 bits . −log2 14 . The nucleotides are monomers of nucleic acids (DNA and RNA) comprising one nitrogenous base, a five-carbon sugar (deoxyribose in DNA and ribose in RNA), and at least one phosphate group. The nitrogenous bases include adenine (A), thymine (T), cytosine (C), guanine (G), and uracil (U). The DNA double helix comprises nucleotides containing the bases A, T, C, and G [44] and maintains its structure due to the complementarity between the nitrogenous bases of each strand of the helix (i.e., the affinity of adenine to thymine and that of cytosine to guanine). As the information in DNA is divided into blocks of nucleotides called genes that possess start and termination sequences, then, we can interpret that biological information is divided into encapsulated data segments. In packet-switching networks, digital information is divided into smaller units known as packets to facilitate processing. Therefore, a packet in a digital network may be analogous to a gene in a biological communication network [32, 35–37]. At the beginning of transcription, the molecular motor RNA polymerase II (the RNAP II enzyme) recognizes a region of the DNA sequence called the promoter region. The promoter hosts the start sequence, which is the location at which RNAP II begins to add nucleotides to create a complementary messenger RNA (mRNA) sequence. During the initiation of transcription, RNAP II produces a complementary single-stranded mRNA copy of one of the two DNA strands. The only difference between RNA and DNA is that RNAP II uses uracil .(U ) instead of thymine .(T ) during this process [32, 35–37]. The enhancer is another important element of the DNA strand that controls the quantity of protein produced according to the amount of mRNA. Because the enhancer controls the amount of information sent to the receiver, this process may be understood as a flow control at the sender end. Data-link layers are responsible for flow control to ensure that a rapid sender cannot swamp a slow receiver with more messages than can be processed. Transcription proceeds specifically unidirectional along one of the DNA strands from the 5’P to the  .3 OH of the deoxyribose phosphate backbone. This specificity is presented also in computer networks in which when the information is transmitted using a serial communication, the less significant bit must be signaled [32, 35–37]. The halt of transcription is accomplished when an appropriate finalization sequence is recognized by RNAP II. In the primary transcript molecule (i.e., premRNA), maturations such as (i) splicing, (ii) capping, and (iii) polyadenylation occur [45]. The information added during capping and polyadenylation may be equivalent to the delimiting data flags used in digital communication systems, for instance, headers and trailers that encapsulate the information (i.e., forms a molecular frame) in the data-link layer in protocol hierarchies in network software. These flags are used for processing and error control; concomitantly, with the utilization of digital flags, the previously mentioned maturations provide stability (control and posterior processing) to the mRNA

186

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

molecules and prevent the degradation of the mRNA by enzymes in the cytosol (intracellular fluid), hence allowing the molecules to advance to the subsequent phases of biological processing. Then, the information added during capping and polyadenylation may be equivalent to a header and trailer that encapsulate the information [32, 35–37]. The mechanical transport of mRNA molecules through the cytosol may be analogous to the transmission of information in wired communications (physical layer task) [32, 35–37]. (b) Translation of DNA The transcription process permitted to copy the biological information from DNA to RNA, this is mandatory because DNA molecules cannot leave the cell’s nucleus. Hence, in this point, the biological DTE must transmit the information to biological DCE (data communication equipment) through a physical interface (as in conventional communication systems). The cytosol may represent the physical interface [32, 35–37]. In a digital communication system, the DCE (codec or modem) is the device required for formatting adequately the information that will be transmitted through a communication channel. The ribosomes and ER (endoplasmic reticulum) represent the biological DCE because through these organelles the genetic information acquires a functional structure (or format when referring to data) that is later released into the biological communication channel and ultimately arrives at the biological receiver. During translation, the “appropriate formatting” of biological data occurs when the information is converted into amino acid chains to obtain functionality inside and outside the cell. Thus, the biological DCE processes (codifies) information via translation and provides a specific input sequence (data in mRNA) that is associated with a specific output (sequence of amino acids); from the digital system’s perspective, this type of codification process corresponds to conventional codification (i.e., a physical layer task). As in every stage of transcription and translation, mother nature has established the opportune intermediate addressing, therefore, the mRNA that leaves the nucleus has an implicit adjacent address that is comparable to a data link layer address to facilitate communication within a direct range of communication; therefore, the mRNA is bound by the ribosomes in the cytosol or those associated with the rough ER (RER). The transmission (movement) of biological information from the nucleus to the ribosomes or ER through the cytosol represents the arriving and moving of information into a biological communication channel, which is considered a task at the physical layer [32, 35–37]. Ribosomes, which are structures that serve as molecular motors, read the information contained in the biological sequence using a codon (i.e., a triplet of nucleotides) system. In the ribosome, the codons in the mRNA are recognized by transfer RNAs (tRNAs) that possess an anticodon (a sequence complementary to a particular codon) associated with a unique amino acid that binds specifically to the molecular structure of that tRNA [32, 35–37].

4.8 Case 8. Gene Expression and Protein Delivery Analysis from a Network. . .

187

Protein synthesis occurs in the ribosome through the signaling of tRNAs that indicate to the ribosomes the beginning (start codon) and end (stop codon) of the process to ensure the proper reading of the biological information. From a digital communication system’s perspective, the analysis of amino acid interactions that form proteins is essential for understanding the evolutionary relationships among organisms, the development of new drugs, and the production of synthetic proteins. From the digital communication systems paradigm, the start and stop codons may correspond to synchronism signals. Synchronism between the source and destination is performed in synchronous transmissions through a start flag. In this type of communication, the transmitter sends the information, and the receiver must collect and process the information. The stop codon in biological signaling may be equivalent to the stop flag in synchronous communications that is used at the destination to indicate the end of the communication. The signals to start and stop the translation of DNA allow the biological clocks present in cells to provide feedback during cellular processes [32, 35–37]. As proteins are generated by the processing of specific amino acids, the quantity of information required to specify one amino acid  from a set of 20 (the 1 total number of amino acids) unambiguously is 4.3 bits . −log2 20 ; then, using 6 bits to define an amino acid signifies an excess of information; nevertheless, this excess information capacity may explain the genetic code’s redundancy [32, 35–37]. In MCs, the modulation (a physical layer task) is functionality at the transmitter to alter the properties of molecules to represent the information that arrives as a concentration of molecules at the receiver. One mechanism for this modulation is to choose one type of molecule from a set of molecule types; this type of modulation is known as MoSK (molecular-shift keying). On the other hand, as genetic code uses 3 nucleotides (codon) to represent a specific amino acid and as there are 20 amino acids, it is necessary to sort the four nucleotides in groups of at least 3 to encode all 20 amino acids in 64 possible combinations (i.e., .43 = 64 because .42 = 16 is not sufficient to encode 20 amino acids). Hence, in [32, 35–37], it is assumed that the biological DCE codifies a set of 6 inputs (3 nucleotides or 6 bits due each one is represented by 2 bits) with a set of 64 outputs (to encode the total of 20 amino acids); and then the DCE in our case may have a 64 MoSK (i.e., there exists 64 different cases to be considered, because for the MoSK modulation scheme, the information is represented by using different types of molecules, for x bits information per symbol, .2x types of molecules are needed to transmit [32, 46]. For having a long nanorange communication (in which case a type of addressing similar to the network address in conventional computer networks is required), in [32, 35–37] is analyzed the production of proteins (e.g., peptide hormones, such as insulin) processed in the ER because many of these types of proteins have a role outside the cell, where they reach the receiver in this called long nanorange. To obtain the mentioned types of proteins is mandatory to tag

188

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

them (i.e., tagging to play a role outside the cell). Thus, a signal recognition particle (SRP) is bound to the amino acid sequence, thereby providing it with an implicit adjacent address via molecular tagging. This molecular tag is comparable to a data-link layer address that facilitates communication within a direct range of communication [32, 35–37]. The principal task of the SRP is to allow the nascent protein to arrive at a channel protein in the ER that oversees the translocation of the protein within the ER. Then, the SRP detaches from the protein and is recycled in the cytosol. Correspondingly, in digital communication systems, after the processing information and control information are used, they are discarded. In this point, inside the ER the proteins are folded and acquire the functional threedimensional structure required for them to accomplish their specific biological functions (equivalent to digital information after processing by the DCE, i.e., having the appropriate format [32, 35–37]. In biological systems, information errors can occur during DNA transcription and translation; similarly, in conventional communication systems, errors can occur in the transmission medium. Errors in cellular processing and information communication are responsible for many medical disorders, such as cancer, autoimmunity, and diabetes [32, 35–37]. (c) The Golgi Apparatus (GA) as an Internet (Border) Router When proteins are functional, the RER transfers the proteins via molecular motors to the GA. Because each protein contains an implicit adjacent address which is comparable to a data-link layer address that facilitates communication within a direct range of communication, the proteins are routed to the appropriate intercellular destinations (i.e., in long nanorange communications with addressing as in network layer); however, the GA determines whether the proteins remain inside the cell [32, 35–37]. During this process, the proteins, along with their information content, move from the RER to the GA, where the information is deposited into vesicles that bind the cis GA face. Then, new vesicles containing the protein information are generated, and other cellular components required for processing the proteins are added. The new vesicles deposit their contents into the medial GA face, and, again, new vesicles containing the protein and the elements necessary for further processing are formed. Finally, the vesicles reach the trans-GA face, where a process identical to the previously defined process occurs; thus, the proteins are inserted into new vesicles but are directed to the endoplasmic membrane to be secreted outside the cell; vesicles may naturally correspond to a molecular frame at the molecular link layer [32, 35–37]. The mentioned functions of the GA are similar to those performed by a border router in a network. A router determines whether the information remains inside the network or leaves the network meanwhile encapsulating and unencapsulated the information through layers in the network. Therefore, the actions of depositing proteins, forming vesicles and attaching information to specify a protein destination, are analogous to those required in the processing of protocol data units (PDUs) in the layers of a router. The layered network

4.8 Case 8. Gene Expression and Protein Delivery Analysis from a Network. . .

189

Fig. 4.9 Layered network communication model of the communication of information from the DTE to the GA

model of the communication of information in DNA transcription and translation and considering the GA as a border router is illustrated in Fig. 4.9. The structure of a layered model decomposes a large-scale system into a set of smaller units (i.e., layers) that are functionally independent of each other and specifies the interactions among the layers. Hence, an advantage of using a stack of layers is the fundamental use of the data-link layer to transform an imperfect channel into a line free of transmission errors or report unsolved problems to the upper layer. Therefore, the application of such a model to biological systems (e.g., drug delivery) could provide high reliability [32, 35–37, 45, 47, 48]. (d) D. Protein delivery through a communication channel. The mode of protein delivery to different destinations in the body is not the same in all cases and depends on the specific requirements of the system involved (e.g., the endocrine system). Hence, we consider cases in which the proteins secreted by the cell (e.g., hormones of a proteinaceous nature) move through the bloodstream (physical transmission medium—active random with drift transport-diffusion with drift) to a target organ (address destination at the network layer) [49]. This type of MC is referred to as intercellular communication (i.e., distances in the range of mm to m). Thus, the movement of molecules in a fluid medium with drift (e.g., the bloodstream) is characterized as follows [32, 35–37, 50]: .

f (t) =



λ 2π t 3

  2 , f or(t) > 0 exp −λ (t−μ) 2 2μ t

(4.5)

where the mean is .μ = d/v, the shape parameter is .λ = d 2 /2D, the velocity of the fluid medium is .v ≥ 0, the diffusion coefficient is D, and the distance from the transmitter to the receiver is d [32, 35–37]. As proteins move through the bloodstream to a target destination, the traffic of biological information from senders to receivers through this media converts

190

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

the blood into shared media. This shared link requires media access control (MAC) to divide it among multiple senders and transmits molecular frames from multiple senders without causing interference between molecular frames. One mechanism for medium access control is time-division multiplexing (TDM), in which different senders transmit molecular frames at different times [32, 35–37]. MC systems and typical communication systems can encounter problems during the transmission of information via communication channels. Specifically, in MC, these problems include biochemical, thermal, and physical noise, interference (which can be controlled by an appropriate transmission rate), and attenuation (which depends on the distance traveled and the physical characteristics of the fluid medium). The noise is any distortion that results in the degradation of the signal at the receiver principally due to its stochastic nature; in MCs, the sources of the noise can be (a) random propagation (diffusion) noise, (b) transmitter emission noise, (c) receiver counting/reception noise, (d) environment noise such as degradation and/or reaction, and (e) multiple transmitters [32, 35–37, 51, 52]. The resulting damage to the signal information can cause latency (i.e., movement delay), which is expressed as .d/v, jitter (i.e., variation in latency) which mathematical expression is D..d/2.v 3 , and the loss rate (i.e., the probability that a molecule transmitted by a biological sender is not received by the intended biological receiver) can increase. The mathematical expression for the loss rate T is .1 − 0 f (t) dt, which assumes that the receiver waits for the time duration T ) [32, 35–37]. Based on the communicational parameters that denote the problems that can occur in a communication channel with noise, the Shannon theorem is used to determine the maximum biological information transfer speed (capacity of a channel) as [32, 35–37] .

C = max {I (X; Y )}

(4.6)

where .I (X; Y ) defines the entropy of the mutual information (MI ) of X and Y . The information signals at the transmission and reception ends are denoted as X and Y , respectively. Figure 4.10 illustrates the result of applying the BlahutArimoto algorithm to maximize the MI . Through this process, the biological transmitter emits a particle in each transmission slot. Also, in Fig. 4.10 the MI (measured in bits) is displayed concerning the diffusion constant and velocity or drift of the fluid. The velocity-diffusion plot can be roughly divided into the following three regions: (i) a diffusion-dominated region in which the MI is relatively insensitive to the velocity; (ii) a high-velocity region in which the MI is insensitive to the diffusion coefficient; and (iii) an intermediate region in which the MI is highly sensitive to both the velocity and diffusion coefficient of the medium. Besides, the MI increases as the velocity increases and reaches its maximum value at .log2 (N ), where N is the number of transmission slots.

4.8 Case 8. Gene Expression and Protein Delivery Analysis from a Network. . .

191

Fig. 4.10 Representation of the MI in relation to the velocity of drift and the diffusion coefficient [53]

If a transmitter does not send any molecule in any transmission slot, then the maximum value of MI will be .log2 (N + 1). It is also evident that when the velocity values are high, the MI values are independent of diffusion coefficient changes [32, 35–37, 53]. (e) Use of information at the Receiver. In the human body (and accordingly to Shannon’s model of communication), the communication of biological information from a transmitter (DTE-DCE) to a receiver (DCE-DTE) is done through the bloodstream (i.e., the communication channel), and in this way a target cell, tissue, or organ performs a physiological function (due to a specific type of biological information which is comparable with network layer addressing); then, we focus in the type of receivers that are located in a long nanorange. Therefore, the transmitter sends the information by using the data stored in the DNA molecules and at the destination can recognize the target cell, tissue, or organ. In terms of the type of proteinaceous hormone involved, the receiver processes the information received. Here, we briefly describe a case in which this processing is performed through ligands and their receptors [32, 35–37, 54]. In nature, signal molecules are received via protein structures called receptors inside the receiver; the receiver consists of a chemical detector, which will sense the concentration of molecules at a specific sample time and demodulate the signal. Therefore, these protein structures can be seen as receiver antennas. Receptors are special protein structures that can bind to specific ligand structures. The binding occurs by intermolecular forces, such as ionic

192

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

Fig. 4.11 Internetwork equivalence of gene expression

bonds, hydrogen bonds, and van der Waals forces. The docking (association) is usually reversible (dissociation). Ligand binding to a receptor alters the receptor’s chemical conformation and the tendency of binding is called affinity. The conformational state of a receptor determines its functional state. Almost all ligand structures in nature capture and remove the information particles from the propagation environment during the detection process [32, 35–37]. In the biological process described, the membrane protein behaves as a transducer that decodes the received signal (i.e., accomplishing the typical DCE tasks at the receiver), triggering several reactions inside the target cell, tissue, or organ of the body, i.e., accomplishing the typical DTE tasks at the receiver). This behavior is comparable to the function performed at the receiver end in digital communication systems to process information that will be profitable at the destination. Once the target cell receives the biological data, the information is communicated to other organelles using an implicit adjacent address that is comparable to the data link layer address used to facilitate communication within a direct range of communication. The received biological message is physically transmitted to the target cell [32, 35–37]. Figure 4.11 presents an end-to-end layered network model of the gene expression of proteins from the typical network perspective of gene expression as an Internet [32, 35–37].

References 1. R. Paridar, N. Mokari, E. Jorswieck, M.R. Javan, On the reception process of molecular communication-based drug delivery (2021). arXiv preprint arXiv:2112.12485 2. L.F. Borges, M.T. Barros, M. Nogueira, Toward reliable intra-body molecular communication: an error control perspective. IEEE Commun. Mag. 59(5), 114–120 (2021) 3. Y. Sun, H. Bian, Y. Chen, A photolysis-assist molecular communication for tumor biosensing. Sensors 22(7), 2495 (2022) 4. D. Jing, Y. Li, A.W. Eckford, An extended Kalman filter for distance estimation and power control in mobile molecular communication (2022). arXiv preprint arXiv:2205.09596

References

193

5. Y. Cevallos, T. Nakano, L. Tello-Oquendo, A. Rushdi, D. Inca, I. Santillán, A.Z. Shirazi, N. Samaniego, A brief review on dna storage, compression, and digitalization. Nano Commun. Netw. 31, 100391 (2022) 6. M. Egan, M. Kuscu, M.T. Barros, M. Booth, A. Llopis-Lorente, M. Magarini, D.P. Martins, M. Schäfer, P. Stano, Toward interdisciplinary synergies in molecular communications: perspectives from synthetic biology, nanotechnology, communications engineering and philosophy of science. Life 13(1), 208 (2023) 7. A. El-atty, M. Saied, R. Bidar, E.-S.M. El-Rabaie, Embedded nano relay for intra-body network-based molecular communications. Wirel. Pers. Commun. 125(4), 3049–3066 (2022) 8. S. Ghavami, R. Adve, F. Lahouti, Information rate analysis of ask-based molecular communication systems with feedback. Nano Commun. Netw. 28, 100339 (2021) 9. Z. Cheng, J. Yan, J. Sun, Y. Tu, K. Chi, Joint optimizations of relays locations and decision threshold for multi-hop diffusive mobile molecular communication with drift. IEEE Trans. NanoBiosci. 21(3), 454–465 (2022) 10. H.K. Rudsari, M. Zoofaghari, M. Veletic, J. Bergsland, I. Balasingham, The end-to-end molecular communication model of extracellular vesicle-based drug delivery (2022). arXiv preprint arXiv:2207.01875 11. M. Femminella, G. Reali, Implementation issues of diffusion-based molecular communications receivers based on transcriptional elements. Digital Signal Process. 124, 103160 (2022) 12. Y. Jian, B. Krishnaswamy, C.M. Austin, A.O. Bicen, J.E. Perdomo, S.C. Patel, I.F. Akyildiz, C.R. Forest, R. Sivakumar, nanons3: simulating bacterial molecular communication based nanonetworks in network simulator 3, in Proceedings of the 3rd ACM International Conference on Nanoscale Computing and Communication (2016), pp. 1–7 13. Y. Jian, B. Krishnaswamy, C.M. Austin, A.O. Bicen, A. Einolghozati, J.E. Perdomo, S.C. Patel, F. Fekri, I.F. Akyildiz, C.R. Forest et al., nanons3: a network simulator for bacterial nanonetworks based on molecular communication. Nano Commun. Netw. 12, 1–11 (2017) 14. Y. Cevallos, L. Tello-Oquendo, D. Inca, D. Ghose, A.Z. Shirazi, G.A. Gomez, Health applications based on molecular communications: a brief review, in 2019 IEEE International Conference on E-health Networking, Application & Services (HealthCom) (IEEE, 2019), pp. 1–6 15. N.A. Ali, M. Abu-Elkheir, Internet of nano-things healthcare applications: requirements, opportunities, and challenges, in 2015 IEEE 11th International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob) (IEEE, 2015), pp. 9–14 16. B.J. Wysocki, T.M. Martin, T.A. Wysocki, A.K. Pannier, Modeling nonviral gene delivery as a macro-to-nano communication system. Nano Commun. Netw. 4(1), 14–22 (2013) 17. M. Dong, W. Li, X. Xu, Evaluation and modeling of HIV based on communication theory in biological systems. Infect. Genet. Evol. 46, 241–247 (2016) 18. U. Chude-Okonkwo, R. Malekian, B. Maharaj, Internet of things for advanced targeted nanomedical applications, in Advanced Targeted Nanomedicine (Springer, Berlin, 2019), pp. 113–124 19. X. Chen, M. Wen, F. Ji, Y. Huang, Y. Tang, A.W. Eckford, Detection interval for diffusion molecular communication: how long is enough (2022)? arXiv preprint arXiv:2204.08636 20. Y. Bayraktar, E. Isik, I. Isik, A. Ozyilmaz, M. Toprak, F. Kahraman Guloglu, S. Aydin, Analyzing of Alzheimer’s disease based on biomedical and socio-economic approach using molecular communication, artificial neural network, and random forest models. Sustainability 14(13), 7901 (2022) 21. C. Fonseca, M.T. Barros, A. Odysseos, S. Kidambi, S. Balasubramaniam, Quasi-spherical absorbing-reactive receiver model of glioblastoma cells for exosome-based molecular communications 22. D. Vimalajeewa, S. Balasubramaniam, D.P. Berry, G. Barry, Virus particle propagation and infectivity along the respiratory tract and a case study for SARS-COV-2. Sci. Rep. 12(1), 1–14 (2022) 23. B.A. Bilgin, E. Dinc, O.B. Akan, Dna-based molecular communications. IEEE Access 6, 73119–73129 (2018)

194

4 Case Studies of Applications of Digital Networks Theories to Molecular. . .

24. A.O. Balghusoon, S. Mahfoudh, Routing protocols for wireless nanosensor networks and internet of nano things: a comprehensive survey. IEEE Access 8, 200724–200748 (2020) 25. O. Yetimoglu, M.K. Avci, B.C. Akdeniz, H.B. Yilmaz, A.E. Pusane, T. Tugcu, Multiple transmitter localization via single receiver in 3-d molecular communication via diffusion. Digital Signal Process. 124, 103185 (2022) 26. Y. Li, L. Lin, W. Guo, D. Zhang, K. Yang, Error performance and mutual information for IoNT interface system. IEEE Internet Things J. 9(12), 9831–9842 (2022) 27. A. Galal, X. Hesselbach, Nano-networks communication architecture: modeling and functions. Nano Commun. Netw. 17, 45–62 (2018) 28. A. El-taweel, S.M. Abd El-atty, S. El-Rabaie, Efficient molecular communication protocol based on mobile ad-hoc nanonetwork. Menoufia J. Electron. Eng. Res. 26(2), 427–443 (2017) 29. F. Walsh, S. Balasubramaniam, D. Botvich, T. Suda, T. Nakano, S.F. Bush, M. Foghlú, Hybrid dna and enzyme based computing for address encoding, link switching and error correction in molecular communication, in International Conference on Nano-Networks (Springer, Berlin, 2008), pp. 28–38 30. IEEE recommended practice for nanoscale and molecular communication framework, in IEEE Std 1906.1-2015 (2016), pp. 1–64 31. IEEE standard data model for nanoscale communication systems, in IEEE Std 1906.1.1-2020 (2020), pp. 1–142 32. Y. Cevallos, T. Nakano, L. Tello-Oquendo, N. Chopra, A.Z. Shirazi, D. Inca, I. Santillán, Theoretical basis for gene expression modeling based on the IEEE 1906.1 standard, in International Conference on Bio-inspired Information and Communication Technologies (Springer, Berlin, 2021), pp. 145–162 33. I.F. Akyildiz, F. Fekri, R. Sivakumar, C.R. Forest, B.K. Hammer, Monaco: fundamentals of molecular nano-communication networks. IEEE Wirel. Commun. 19(5), 12–18 (2012) 34. M. Egan, B.C. Akdeniz, B.Q. Tang, Stochastic reaction and diffusion systems in molecular communications: recent results and open problems. Digital Signal Process. 124, 103117 (2022) 35. Y. Cevallos, L. Molina, A. Santillán, F. De Rango, A. Rushdi, J.B. Alonso, A digital communication analysis of gene expression of proteins in biological systems: a layered network model view. Cogn. Comput. 9(1), 43–67 (2017) 36. Y. Cevallos, L. Tello-Oquendo, D. Inca, C. Palacios, L. Rentería, Genetic expression in biological systems: a digital communication perspective. Open Bioinform. J. 12(1) (2019) 37. Y. Cevallos, T. Nakano, L. Tello-Oquendo, D. Inca, I. Santillán, A.Z. Shirazi, A. Rushdi, N. Samaniego, Modeling gene expression and protein delivery as an end-to-end digital communication system. Open Bioinform. J. 14(1) (2021) 38. I. Iqbal, M. Nazir, A. Sabah, Design of energy-efficient protocol stack for nanocommunication using greedy algorithms. J. Comput. Netw. Commun. 2022 (2022) 39. M.J. Salariseddigh, U. Pereg, H. Boche, C. Deppe, V. Jamali, R. Schober, Deterministic identification for molecular communications over the poisson channel (2022). arXiv preprint arXiv:2203.02784 40. A.D. Odysseos, T. Nakano, Bionanomachine diagnostics and nanonetwork therapeutic in brain malignancies with bionanodevice interfaces. IEEE Trans. Mol. Biol. Multi-Scale Commun. 8(1), 28–35 (2021) 41. A. Awad, M.M. Fouda, M.M. Khashaba, E.R. Mohamed, K.M. Hosny, Utilization of mobile edge computing on the internet of medical things: a survey. ICT Express (2022) 42. J.T. Gómez, K. Pitke, L. Stratmann, F. Dressler, Age of information in molecular communication channels. Digital Signal Process. 124, 103108 (2022) 43. W. Pan, X. Chen, X. Yang, N. Zhao, L. Meng, F.H. Shah, A molecular communication platform based on body area nanonetwork. Nanomaterials 12(4), 722 (2022) 44. L. Felicetti, M. Femminella, G. Reali, A simple and scalable receiver model in molecular communication systems, in Proceedings of the 3rd ACM International Conference on Nanoscale Computing and Communication (2016), pp. 1–2 45. S. Ahmed, J. Hu, S.M. Naqvi, Y. Zhang, L. Linze, A.M. Iderawumi, Molecular communication network and its applications in crop sciences. Planta 255(6), 1–12 (2022)

References

195

46. Y. Lu, Error correction codes for molecular communication systems, Ph.D. Dissertation, University of Warwick, 2016 47. Z. Cheng, J. Sun, J. Yan, Y. Tu, Optimizations for mobile mimo relay molecular communication via diffusion with network coding. KSII Trans. Internet Inform. Syst. 16(4), 1373–1391 (2022) 48. I. Isik, How mobility of transmitter and receiver affects the communication quality. AIP Adv. 12(2), 025205 (2022) 49. L. Chouhan, P.K. Upadhyay, P.K. Sharma, A.M. Salhab, On anomalous diffusion of devices in molecular communication network (2022). arXiv preprint arXiv:2203.14808 50. S. Sarkar, S. Choubey, Efficacy of information transmission in cellular communication (2020). arXiv preprint arXiv:2010.05291 51. S. Figueiredo, N. Souto, F. Cercas, Low-complexity channel codes for reliable molecular communication via diffusion. Sensors 22(1), 41 (2021) 52. F. Vakilipoor, F. Ratti, H. Awan, M. Magarini, Low complexity receiver design for time-varying poisson molecular communication channels with memory. Digital Signal Process. 124, 103187 (2022) 53. P. Raut, N. Sarwade, Establishing a molecular communication channel for nano networks. Int. J. VLSI Design Commun. Syst. 4(2), 27 (2013) 54. M. Kuscu, O.B. Akan, Detection in molecular communications with ligand receptors under molecular interference. Digital Signal Process. 124, 103186 (2022)

Index

B Biological communications, 131, 173, 185, 186 Biological communications applications, 173

C Codification and modulation in molecular physical channels, 107–109, 186 Communications models, 6, 8, 14, 16, 39, 155, 182

I Information and communication technology techniques, 3, 4, 69, 184 Interference and attenuation in molecular physical channels, 51 Internetwork reference models, 14, 184

L Link and physical layer’s applications, 13–56, 63–158 Link layers, 13–56

M Medical applications, 35, 63 Molecular application, 18–20 Molecular communications network architecture, 14–56 Molecular communications systems, 69–158 Molecular communications systems and components, 71–81 Molecular physical layer, 150–158

N Network, 1, 13, 63, 168 Networks and components, 13, 65 Noise, 4, 8, 49, 69, 72, 77, 79, 80, 86–95, 98–100, 104, 125, 129, 138, 150, 155, 173, 190

P Propagation models, 150

T Transport, 1, 13–56, 70–72, 75, 78, 79, 84, 102, 132, 158, 170–173, 177, 180, 182, 186, 189

© The Editor(s) (if applicable) and The Author(s),under exclusive license to Springer Nature Switzerland AG 2024 Y. Cevallos et al., Molecular Communications, https://doi.org/10.1007/978-3-031-36882-0

197