Energy Harvesting Communications E-Book

Yunfei Chen

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Serie: Wiley - IEEE
• Sprache: Englisch

Seitenzahl: 537

Veröffentlichungsjahr: 2018

Beschreibung

Provides a systematic overview of a hot research area, examining the principles and theories of energy harvesting communications This book provides a detailed and advanced level introduction to the fundamentals of energy harvesting techniques and their use in state-of-the-art communications systems. It fills the gap in the market by covering both basic techniques in energy harvesting and advanced topics in wireless communications. More importantly, it discusses the application of energy harvesting in communications systems to give readers at different levels a full understanding of these most recent advances in communications technologies. The first half of Energy Harvesting Communications: Principles and Theories focuses on the challenges brought by energy harvesting in communications. The second part of the book looks at different communications applications enhanced by energy harvesting. It offers in-depth chapters that: discuss different energy sources harvested for communications; examine the energy harvesters used for widely used sources; study the physical layer and upper layer of the energy harvesting communications device; and investigate wireless powered communications, energy harvesting cognitive radios, and energy harvesting relaying as applications. * Methodically examines the state-of-the-art of energy harvesting techniques * Provides comprehensive coverage from basic energy harvesting sources and devices to the end users of these sources and devices * Looks at the fundamental principles of energy harvesting communications, and biomedical application and intra-body communications * Written in a linear order so that beginners can learn the subject and experienced users can attain a broader view Written by a renowned expert in the field, Energy Harvesting Communications: Principles and Theories is an excellent resource for students, researchers, and others interested in the subject.

Leseprobe

Table of Contents

Cover

Preface

Acronyms

1 Introduction

1.1 Background

1.2 Relationship with Green Communications

1.3 Potential Applications

1.4 Outline of Chapters

2 Energy Sources

2.1 Introduction

2.2 Types of Sources

2.3 Predictive Models of Sources

2.4 Summary

3 Energy Harvesters

3.1 Introduction

3.2 Photovoltaic Panels

3.3 Radio Frequency Energy Harvester

3.4 Overall Models

3.5 Battery and Supercapacitor

3.6 Summary

4 Physical Layer Techniques

4.1 Introduction

4.2 Effect of Energy Harvesting

4.3 Energy Harvesting Detection

4.4 Energy Harvesting Estimation

4.5 Energy Transmission Waveform

4.6 Other Issues and Techniques

4.7 Summary

5 Upper Layer Techniques

5.1 Introduction

5.2 Media Access Control Protocols

5.3 Routing Protocols

5.4 Other Issues in the Upper Layers

5.5 Summary

6 Wireless Powered Communications

6.1 Introduction

6.2 Types of Wireless Powered Communications

6.3 Simultaneous Wireless Information and Power Transfer

6.4 Hybrid Access Point

6.5 Power Beacon

6.6 Other Issues

6.7 An Example: Wireless Powered Sensor Networks

6.8 Summary

7 Energy Harvesting Cognitive Radios

7.1 Introduction

7.2 Conventional Cognitive Radio

7.3 Types of Energy Harvesting Cognitive Radio

7.4 From the Secondary Base Station

7.5 From the Primary User

7.6 From the Ambient Environment

7.7 Information Energy Cooperation

7.8 Other Important Issues

7.9 Summary

8 Energy Harvesting Relaying

8.1 Introduction

8.2 Conventional Relaying

8.3 Types of Energy Harvesting Relaying

8.4 From the Ambient Environment

8.5 From the Power Transmitter

8.6 From the Source

8.7 Other Important Issues

8.8 Summary

References

Index

End User License Agreement

List of Tables

Chapter 2

Table 2.1 Typical amount of energy from different sources.

Chapter 3

Table 3.1 Fitted parameters for some energy harvesters.

Chapter 7

Table 7.1 CR systems with different energy sources for PU and CR.

Table 7.2 List of channel gains.

Chapter 8

Table 8.1 Energy harvesting relaying systems with different energy sources.

List of Illustrations

Chapter 1

Figure 1.1 Relationship between energy harvesting communications and green comm...

Figure 1.2 Organization of chapters in the book.

Chapter 2

Figure 2.1 Some commonly used energy sources for energy harvesting wireless com...

Figure 2.2 Example measurements of ambient RF energy at different time instants ...

Figure 2.3 Time series of measured power in the 3G band from 1805 to 1880 MHz in...

Figure 2.4 Comparison of prediction errors for different machine learning algori...

Figure 2.5 Effect of prediction error on wasted and missed energy using linear r...

Figure 2.6 Comparison of predicted and measured power using the wavelets method ...

Chapter 3

Figure 3.1 Diagram of a typical PV cell.

Figure 3.2 Comparison of three one‐diode models.

Figure 3.3 Comparison of three two‐diode models.

Figure 3.4 Diagram of a typical RF energy harvester.

Figure 3.5 Three different matching networks: (a) transformer; (b) parallel coil...

Figure 3.6 Comparison of the fitting methods in 3.22 and 3.23 with the experimen...

Figure 3.7 Comparison of the Fourier model and the experimental value for freque...

Figure 3.8 Comparison of the Fourier model and the experimental value for distan...

Figure 3.9 Comparison of the actual power and predicted power using linear regre...

Figure 3.10 Two‐state Markov chain model.

Chapter 4

Figure 4.1 Diagrams of (a) OSI and (b) TCP/IP models.

Figure 4.2

versus

in Rayleigh fading channels when

dB.

Figure 4.3

versus

in Rayleigh fading channels when

dB.

Figure 4.4

versus

in Rician fading channels when

dB and

.

Figure 4.5

versus

in Rayleigh fading channels when

dB.

Figure 4.6

versus

in Rayleigh fading channels when

dB.

Figure 4.7 Average BER versus

in Rayleigh fading channels.

Figure 4.8 Average BER versus

in Rayleigh fading channels.

Figure 4.9 Comparison of conventional BFSK and energy harvesting BFSK in Rayleig...

Figure 4.10 Average achievable rate (bits/s/Hz) versus

in Rayleigh fading chan...

Figure 4.11 Average achievable rate (bits/s/Hz) versus

in Rayleigh fading chan...

Figure 4.12 Comparison of average achievable rates (bits/s/Hz) for conventional ...

Figure 4.13 Comparison of conventional and energy harvesting communications usin...

Figure 4.14 Comparison of conventional and energy harvesting communications usin...

Figure 4.15 MSE of

and

versus

for different values of SNR in Scheme 1.

Figure 4.16 MSE of

and

versus

for different values of SNR in Scheme 1.

Figure 4.17 MSE of

and

versus

for different values of SNR in Scheme 2.

Figure 4.18 MSE of

and

versus

for different values of SNR in Scheme 2.

Figure 4.19 MSE of

and

versus

for different values of SNR in Scheme 3.

Figure 4.20 MSE of

and

versus

for different values of SNR in Scheme 3.

Figure 4.21 MSE of

and

versus

for different values of SNR in Scheme 4.

Figure 4.22 MSE of

and

versus

for different values of SNR in Scheme 4.

Figure 4.23 MSE of

versus

for different schemes.

Figure 4.24 MSE of

versus

for different schemes.

Figure 4.25 The considered data frame with three parts.

Figure 4.26

versus

.

Figure 4.27 A diagram of the circuit of the remote device.

Figure 4.28 The data packet structure that splits pilots.

Figure 4.29 The data packet structure that splits data.

Figure 4.30 The data packet structure that splits both pilots and data.

Figure 4.31 Outage versus

.

Figure 4.32Figure 4.32 Outage versus

.

Figure 4.33Figure 4.33 BER versus

.

Figure 4.34 BER versus

.

Chapter 5

Figure 5.1 The state transition process of each node.

Figure 5.2 The transmission scheduling problem.

Figure 5.3 The link layer channel model.

Chapter 6

Figure 6.1 Some scenarios where wireless power is a promising alternative.

Figure 6.2 Different types of wireless powered communications: (a) SWIPT; (b) HA...

Figure 6.3

versus

in 6.2.

Figure 6.4 Comparison of (a) TS and (b) PS.

Figure 6.5 The rate‐energy function for TS.

Figure 6.6 The rate‐energy function for PS.

Figure 6.7 A unified structure for TS and PS.

Figure 6.8 HAP wireless powered communications: (a) HAP structure; and (b) HAP l...

Figure 6.9

versus

when there is one node.

Figure 6.10 Average rate versus

for different values of

in the Jakes' model.

Figure 6.11 Average rate versus

for different values of

in the channel.

Figure 6.12 Average rate versus

for different values of the Nakagami‐

paramet...

Figure 6.13 Maximum average rate versus

for different values of

in the Jakes...

Figure 6.14 Optimum

versus

for different values of

in the Jakes' model.

Figure 6.15 Maximum average rate versus

for different values of the Nakagami‐

Figure 6.16 Optimum

versus

for different values of the Nakagami‐

parameter ...

Figure 6.17 Average BER versus

for different values of

in the Jakes' model.

Figure 6.18Figure 6.18 Average BER versus

for different values of

in the cha...

Figure 6.19 Average BER versus

for different values of the Nakagami‐

paramete...

Figure 6.20 Average sum rate versus

for two devices.

Figure 6.21 A mobile receiver used in PB‐based wireless powered communications.

Figure 6.22 Average SINR versus

for different values of

,

and

.

Figure 6.23 Average rate versus

for different values of

,

and

.

Figure 6.24 Rate of HAP versus

for different approximations.

Figure 6.25 Rate versus

for different wireless powered networks when

and

f...

Figure 6.26 Rate versus

for different wireless powered networks when

and

f...

Figure 6.27 Comparison of PB using different

and HAP.

Figure 6.28 Comparison of different

for HAP.

Figure 6.29

versus

for different topologies when only one node transmits in ...

Figure 6.30

versus

for different topologies when only one node transmits in ...

Figure 6.31

versus

for different topologies when two nodes transmit data sim...

Figure 6.32 The average value of

for different topologies when two nodes trans...

Figure 6.33 Comparison of different transmission strategies.

Chapter 7

Figure 7.1 CR systems using (a) temporal and (b) spatial opportunities.

Figure 7.2 The frame structure of a conventional CR.

Figure 7.3 Comparison of (a) conventional CR and (b) energy harvesting CR.

Figure 7.4 Comparison of interweave CR and underlay CR when the temporal opportu...

Figure 7.5 Diagram of an overlay CR system.

Figure 7.6 ROC of the energy detector in 7.4.

Figure 7.7 Comparison of different feature detectors.

Figure 7.8 Performances of different feature detectors.

Figure 7.9 Comparison of (a) “harvest‐use” and (b) “harvest‐store‐use”.

Figure 7.10 CR system that has its secondary base station as the energy source.

Figure 7.11

versus

.

Figure 7.12 CR system that harvests energy from the PU.

Figure 7.13 The optimum information rate in 7.62 versus

.

Figure 7.14 The optimum information rate in 7.67 versus

.

Figure 7.15 Diagram of energy harvesting CR where the PU uses wireless power.

Figure 7.16

versus

.

Figure 7.17

versus

.

Figure 7.18 CR system that harvests energy from the ambient environment.

, dete...

Chapter 8

Figure 8.1 A relaying system for diversity gain.

Figure 8.2 A relaying system for coverage extension.

Figure 8.3 Accuracy of the equivalent end‐to‐end SNR for the BER of DF.

Figure 8.4 Comparison of ergodic capacities for AF and DF.

Figure 8.5 Different relay selection schemes when

and

.

Figure 8.6 Different relay selection schemes when

and

.

Figure 8.7 Diagram of two‐way relaying.

Figure 8.8 Both source and relay harvest energy from the ambient environment.

Figure 8.9 Both source and relay harvest from a dedicated power transmitter with...

Figure 8.10

versus

.

Figure 8.11 Both source and relay harvest from a dedicated power transmitter wit...

Figure 8.12 Comparison of (a) 8.83 and (b) 8.84 for the case of two users.

Figure 8.13 The relay harvests energy from the source.

Figure 8.14 (a)

and (b)

versus

for TS AF.

Figure 8.15 (a)

and (b)

versus

for PS AF.

Figure 8.16 Throughput versus

for different values of

,

and

using instant...

Figure 8.17 BER versus

for different values of

,

and

using instantaneous ...

Figure 8.18 Optimum BER versus

or

when

using instantaneous transmission.

Figure 8.19 The optimum

achieving minimum BER versus

or

when

using insta...

Figure 8.20 Throughput versus

for different values of

using delay‐tolerant t...

Figure 8.21 BER versus

for different values of

using error‐tolerant transmis...

Figure 8.22 Throughput versus

for different values of

using delay‐constraine...

Figure 8.23 BCR versus

for different values of

using error‐constrained trans...

Figure 8.24 Performance gain versus

when

m,

m,

and

for variable‐gain A...

Figure 8.25 Performance gain versus

when

W,

m,

m, and

for variable‐gai...

Figure 8.26 Performance gain versus

when

W,

,

, and

for variable‐gain AF...

Figure 8.27 Performance gain versus

when

W,

m,

m, and

for variable‐gai...

Figure 8.28 Performance gain versus

when

m,

m,

, and

for fixed‐gain AF,...

Figure 8.29 Performance gain versus

when

W,

m,

m, and

for fixed‐gain A...

Figure 8.30 Performance gain versus

when

W,

,

, and

for fixed‐gain AF, w...

Figure 8.31 Performance gain versus

when

W,

m,

m, and

for fixed‐gain A...

Figure 8.32

versus

for different values of (a)

and (b)

in TS.

Figure 8.33

versus

for different values of (a)

and (b)

parameters in TS.

Figure 8.34

versus

for different values of (a)

and (b)

parameters in PS.

Figure 8.35

versus

for different values of (a)

and (b)

parameters in PS.

Figure 8.36 The values of

or

versus hop number when

dB,

, and

.

Figure 8.37

for different values of

,

and

for fixed‐gain AF with TS.

Figure 8.38Figure 8.38

for different values of

,

and

for variable‐gain AF...

Figure 8.39

for different values of

,

and

for DF with TS.

Figure 8.40

for different values of

,

and

for fixed‐gain AF with PS.

Figure 8.41

for different values of

,

and

for fixed‐gain AF with TS.

Figure 8.42

for different values of

,

and

for fixed‐gain AF with PS.

Figure 8.43 Values of

or

versus hop number.

Figure 8.44 Largest number of hops versus

.

Guide

Cover

Table of Contents

Begin Reading

Pages

iii

xi

xii

xiii

xiv

1

2

3

4

5

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

Energy Harvesting Communications

Principles and Theories

Yunfei Chen

University ofWarwick Coventry, UK

Copyright

This edition first published 2019

© 2019 John Wiley & Sons Ltd

Edition History

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.

The right of Yunfei Chen to be identified as the author of this work has been asserted in accordance with law.

Registered Offices

John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA

John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

Editorial Office

The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.

Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print versions of this book may not be available in other formats.

Limit of Liability/Disclaimer of Warranty

While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

Library of Congress Cataloging‐in‐Publication Data

Names: Chen, Yunfei, 1976‐ author.

Title: Energy harvesting communications : principles and theories / Yunfei Chen, University of Warwick, Coventry, UK.

Description: First edition. | Hoboken, NJ : John Wiley & Sons, Inc., [2019] | Includes bibliographical references and index. |

Identifiers: LCCN 2018038943 (print) | LCCN 2018039730 (ebook) | ISBN 9781119383055 (Adobe PDF) | ISBN 9781119383086 (ePub) | ISBN 9781119383000 (hardcover)

Subjects: LCSH: Wireless communication systems–Power supply. | Energy harvesting. | Microharvesters (Electronics)

Classification: LCC TK5103.17 (ebook) | LCC TK5103.17 .C44 2019 (print) | DDC 621.382/32–dc23

LC record available at https://lccn.loc.gov/2018038943

Cover Design: Wiley

Cover Images: © Verticalarray/Shutterstock; © Iscatel/Shutterstock

Dedication

To my parents

Preface

Wireless communication has provided unprecedented convenience for our daily lives over the past few decades. This convenience largely comes from the replacement of data cables with wireless interconnections. However, the remaining power cables or batteries used in conventional wireless systems are still restricting their mobility or lifetime and hence, are preventing them from being deployed in more and wider applications. Meanwhile, wireless power has emerged as a recent innovation to substitute the power cable. Many breakthroughs have been made. The innovation is further boosted by green communications that aims to meet the governmental targets for emission reduction by harvesting solar energy, wind energy, and other renewable sources.

Given these recent development, it is reasonable to adopt wireless power or energy harvesting in communications so that the last cable in wireless systems can be removed to exploit the full potential of wireless communications. This leads to energy harvesting wireless communications, which is the topic and the motivation of this book.

The use of harvested energy brings big challenges to system designs. First and most importantly, the energy source becomes random or dynamic. This leads to fundamental changes to wireless system designs. Secondly, energy harvesting changes the characteristics of signals and channels utilized in system designs. On the other hand, the use of harvested energy also creates great opportunities. It allows perpetual and sustainable operations of wireless systems. Many conventional wireless systems can be upgraded by adding the energy harvesting functionality to improve their sustainability. For example, sensor networks can use energy harvesting to prolong their lifetime. Cognitive radios can exchange energy for transmission opportunities. Relaying networks can encourage more idle nodes to be involved in relaying by offering them wireless power.

Chapters 1–5 focus on the challenges brought by energy harvesting in wireless communications. Chapters 6–8 focus on different wireless applications enhanced by energy harvesting. Specifically, this book will follow the flow of energy from the energy source, to the energy harvester, to the wireless device, and then to the application. Chapter 1 gives a brief introduction of energy harvesting wireless communications. Chapter 2 discusses different energy sources harvested for wireless communications. Chapter 3 examines the energy harvesters used for two widely used sources. Chapters 4 and 5 study the physical layer and upper layer of the energy harvesting wireless device, respectively. Chapters 6–8 investigate wireless powered communications, energy harvesting cognitive radios and energy harvesting relaying as applications. The whole book focuses on principles and theories rather than systems and implementations.

I would like to acknowledge the contributions made by my students, Freeha Azmat, Youwen Wang, and Idris Adenopo, on the modeling parts, and the useful discussions with my visitors, Yan Gao and Zhibin Xie, on energy harvesting cognitive radios and interference. I am also very grateful to some of my collaborators, Professor Ning Cao, Dr Nan Zhao, Dr Hee‐Seok Oh, Dr Gaofeng Pan, Dr Wei Feng, Dr Zhutian Yang, and Dr Mohamed‐Slim Alouini, for their support. I would like to sincerely thank my PhD adviser, Dr Norman C. Beaulieu, who showed me the path to an academic career. Last but not least, I would like to give a special thank you to my wife for supporting me through out this book project.

I apologize in advance for any errors that may have occurred and welcome any comments and suggestions for further improvements.

Yunfei Chen

Coventry, UK, May 2018

Acronyms

3G

Third Generation

4G

Fourth Generation

5G

Fifth Generation

AC

Alternating Current

AF

Amplify‐and‐Forward

AP

Access Point

AWGN

Additive White Gaussian Noise

BCR

Bit Correct Rate

BER

Bit Error Rate

BFSK

Binary Frequency Shift Keying

BPSK

Binary Phase Shift Keying

CDF

Cumulative Distribution Function

CDMA

Code Division Multiple Access

CR

Cognitive Radio

CSMA/CA

Carrier Sensing Multiple Access with Collision Avoidance

D2D

Device‐to‐Device

DF

Decode‐and‐Forward

DPS

Dynamic Power Splitting

DPSK

Differential Phase Shift Keying

EA‐MAC

Energy Adaptive Media Access Control

EME

Average Energy to Minimum Eigenvalue

EWMA

Exponentially Weighted Moving Average

FCC

Federal Communications Commission

GSM

Global System for Mobile Communications

HAP

Hybrid Access Point

HE‐MAC

Harvest‐then‐Transmit Media Access Control

LOS

Line‐Of‐Sight

MAC

Media Access Control

MB

Moment‐Based

ME

Maximum Eigenvalue

MIMO

Multiple‐Input‐Multiple‐Output

MME

Maximum‐to‐Minimum Eigenvalue

ML

Maximum Likelihood

MSE

Mean Squared Error

NLOS

Non‐Line‐Of‐Sight

NOMA

Non‐Orthogonal Multiple Access

NRMSE

Normalized Root Mean Squared Error

OFDM

Orthogonal‐Frequency‐Division‐Multiplexing

OFDMA

Orthogonal Frequency Division Multiple Access

OMA

Orthogonal Multiple Access

OSI

Open Systems Interconnection

PAPR

Peak‐to‐Average‐Power Ratio

PB

Power Beacon

PDF

Probability Density Function

PHY

Physical

PS

Power Splitting

PSK

Phase Shift Keying

PU

Primary User

PV

Photovoltaic

QoS

Quality of Service

RC

Resistor‐Capacitor

RD

Relay to Destination

RF

Radio Frequency

RFID

Radio Frequency Identification

ROC

Receiver Operating Characteristics

SIR

Signal‐to‐Interference Ratio

SINR

Signal‐to‐Interference‐plus‐Noise Ratio

SNR

Signal‐to‐Noise Ratio

SR

Source to Relay

STC

Standard Testing Conditions

SWIPT

Simultaneous Wireless Information and Power Transfer

TCP/IP

Transmission Control Protocol/Internet Protocol

TDD

Time‐Division‐Duplex

TDMA

Time Division Multiple Access

TS

Time Switching

QoS

Quality of Service

WCMA

Weather Conditioned Moving Average

WPC

Wireless Powered Communications

WSN

Wireless Sensor Network

1Introduction

1.1 Background

Energy harvesting wireless communication is one of the most recent advances in communications techniques. It refers to any communications systems that use devices powered by energy from either the ambient environment or a dedicated power transmitter in a cable‐less or battery‐free way. This new method of communications has two main benefits. First, conventional communications devices rely either on batteries or fixed mains connections for energy supply. However, all batteries have a limited lifetime, while mains connections are not flexible. Energy harvesting provides a promising solution to perpetual and flexible operations of communications devices. Specifically, wireless communication replaces the data cable with wireless interconnection, while energy harvesting aims to replace the power cable with harvested energy, the very last cable in wireless communications. Together, energy harvesting wireless communication provides unprecedented convenience for our daily life. Secondly, energy efficiency is a key issue in wireless communications systems. This is particularly important for conventional wireless devices that rely on batteries, such as sensor nodes or mobile phones. Due to this importance, many studies have been conducted to improve the energy efficiency of wireless communications, including green communications. In fact, most studies on modern communications systems are about their energy efficiency or spectral efficiency. For energy harvesting wireless communications, since the devices are powered by the ambient environment or power transmitter, this problem is less severe for the communication devices. Thus, the main benefits of energy harvesting are the increased convenience and the improved energy efficiency in communications.

These two benefits are achieved at certain costs and have generated several issues. In energy harvesting wireless communications, although the energy supply becomes wireless and endless to provide convenience and efficiency, respectively, the quality of the energy supply drops. For the energy harvested from the ambient environment, such as the sun or wind, there is great uncertainty in the energy availability. This is due to the unpredictable or uncontrollable changes in the ambient environment. For the energy harvested from the power transmitter, there is less uncertainty but to allow energy harvesting and information delivery at the same device, one has to either use two separate sets of equipment or share the same set of equipment. If energy harvesting and information delivery use separate equipment, the hardware cost increases. For example, two radio frequency (RF) fronts may be needed. If both share the same equipment, the coordination between energy harvesting and information delivery becomes complicated. The system throughput may also decrease.

These issues have significant impact on wireless system designs, as power and bandwidth are two precious resources for communications. Hence, most research works in energy harvesting wireless communications focus on these two issues. For example, when the energy supply is insufficient, the wireless node may not be able to transmit or receive data when it wants to. Also, when energy harvesting and information delivery are performed in the same frequency band using the same transceiver, the original time interval for data transmission may have to be divided into two parts for the best coordination between energy harvesting and information delivery. In the physical layer, these will affect signal transmission, signal detection, and signal estimation. In the upper layer, these will affect user scheduling, channel assignment, message control, and message routing. These issues will also change the ways in which many recently proposed wireless systems operate. Thus, the effect of energy harvesting and new designs based on energy harvesting will be investigated for both legacy systems and recently proposed systems in this book.

1.2 Relationship with Green Communications

Green communications is a concept proposed recently to tackle the energy efficiency problem of communications as well as to reduce the CO2 footprint of communications devices. According to the report from the International Telecommunication Union, information and communications technologies account for 2% of all CO2 emissions. Among them, mobile calls contribute 125 million tonnes of CO2 emissions every year. These figures are steadily increasing due to the fact that more and more communications systems are being deployed. Thus, under the pressure from governments to achieve emission reduction targets by 2020 and 2050, telecommunications carriers will have to take action. Moreover, this action will eventually reduce their energy bills as well. For example, by shutting down or properly scheduling base stations in the evenings when the traffic is low, one of the methods proposed for green communications, the carriers can reduce their operational costs.

Energy harvesting can be one way of implementing green communications by powering wireless devices using environmentally friendly energy sources, such as solar power and wind power. This can replace or reduce the battery power and the mains power to save energy. However, energy harvesting does not necessarily save energy. For example, in wireless energy harvesting using power transmitters, a large amount of energy will be wasted during the transmission loss in order to provide convenience at the remote node. Thus, energy harvesting communications and green communications are related to each other but do not belong to each other. Figure 1.1 shows the relationship between energy harvesting wireless communications and green communications.

Figure 1.1 Relationship between energy harvesting communications and green communications.

1.3 Potential Applications

Energy harvesting wireless communication has a lot of interesting applications. One main application is wireless sensor networks (Sudevalayam and Kulkarni 2011). Many sensor nodes are designed for low‐power and low‐data‐rate scenarios, which are very suitable for energy harvesting communications. These applications mainly use energy harvested from the ambient energy sources, such as the sun and the wind. Thus, they also fall into the category of green communications. In other applications, wireless energy harvesting from a dedicated wireless power transmitter can also be used, such as cellular communications (Huang and Lau 2014). Among these applications, the fifth‐generation (5G) mobile communications system is a good use case.

1.3.1 Energy Harvesting for 5G

It is noted that 5G is actually a general backbone network that aims to support the Internet of Things, vehicular communications and other applications, in addition to cellular communications. In this sense, it is an enabling technology for wireless sensor networks too. In this section, we mainly focus on its cellular communications function.

For example, in Liu et al. (2015c), an integrated energy and spectrum harvesting mechanism for 5G networks has been proposed. Spectrum harvesting refers to cognitive radio operations for spectrum opportunities, while energy harvesting uses the ambient energy opportunities to support short‐distance communications. A multi‐tier network was considered and the effects of spectrum and energy harvesting on device‐to‐device (D2D) communications, Femtocell, Picocell and Macrocell operations have been discussed. It was shown that the aggregate network throughput of such a network can be greatly improved due to spectrum and energy harvesting.

In Hossain and Hasan (2015), a general overview on the 5G cellular network was given. Several key enabling technologies and research challenges have been discussed. Among them, the importance of using energy harvesting to improve the energy efficiency of 5G systems has been investigated. It was suggested that, for 5G services that do not have strict requirements on reliability or quality of service (QoS), ambient sources can be used for harvesting, similar to Liu et al. (2015a). However, for 5G services that require QoS, dedicated RF power transmitters can be used so that energy is always available when needed. A similar discussion is found in Buzzi et al. (2016) with more recent reviews.

In Ding et al. (2017b), energy harvesting was combined with non‐orthogonal multiple access (NOMA) to improve energy efficiency and spectrum efficiency in 5G networks. In particular, the NOMA users could be powered by energy harvesting to relay the information in 5G networks. Similarly, in Khan et al. (2016), energy harvesting was combined with millimeter waves, another important 5G technique. In this case, 5G devices tried to harvest energy from the millimeter waves.

There are many other studies on the application of energy harvesting in 5G mobile communications. Since this is not the focus of the book, we have only given a very brief discussion here. Also, owing to the importance of energy harvesting, there is a huge investment in the integration of energy harvesting into existing systems around the world. Due to the enormous number of projects funded in this area, we do not discuss them here.

1.4 Outline of Chapters

In the following chapters, different aspects of energy harvesting wireless communications will be discussed. In Chapter 2, we will discuss different sources of energy that can be harvested and used for communications. Empirical and mathematical models will be examined. This deals with the energy source in energy harvesting wireless communications. In Chapter 3, we will study the relevant energy harvesters for different sources. Their principles and theories will be discussed. This deals with the energy conversion in energy harvesting wireless communications. In Chapter 4and Chapter 5, the effect of energy harvesting and new techniques based on energy harvesting will be investigated for the physical layer and the upper layer at the wireless device, respectively. This deals with the energy usage in energy harvesting wireless communications. Finally, from Chapter 6 to Chapter 8, the application of energy harvesting in recently proposed systems, including wireless powered systems, cognitive radio systems and relaying systems, will be studied. These deal with the application of energy harvesting wireless communications. Figure 1.2 shows a diagram of how these chapters are organized, related, and what kind of problems they deal with.

Figure 1.2 Organization of chapters in the book.

2Energy Sources

2.1 Introduction

There are many different types of energy sources available for harvesting in wireless communications. Depending on their characteristics, they can be categorized as follows:

Uncontrollable and unpredictable

: These energy sources cannot be controlled to generate the amount of energy required at a specific time in a specific location. Also, they do not follow commonly used predictive models or implementation of such predictive models is too complicated for relevant applications. An example of such an energy source is mechanical vibration. A piezoelectric or electrostatic energy harvester can convert the vibrational energy into electricity but it may be hard to predict when or where the vibration will occur and it is even harder to generate it intentionally (Mitcheson et al.

2008

).

Uncontrollable but predictable

: These energy sources cannot be controlled to generate the energy when and where it is needed. However, their generation follows certain patterns that have been well studied and are relatively predictable with acceptable errors. For example, solar energy is mainly determined by solar activities and weather conditions. It is hard to control the movement of the sun or the weather to achieve the level of solar energy desired but solar energy has strong diurnal and seasonal cycles that can be predicted (Bergozini et al.

2010

). This prediction can be further improved by incorporating weather data in the forecast.

Controllable and partially predictable

: These energy sources can be controlled to produce the amount of energy required at a specific time in a specific location by the wireless device. In other words, these energy sources are controlled by the communications system. Wireless power is a good example of energy sources in this category. In wireless powered communications systems, a radio frequency (RF) signal can be sent by the power transmitter to the remote wireless device for electricity. Also, in an indoor environment, the indoor light can be controlled for the wireless device to harvest its energy using a photovoltaic cell (Wang et al.

2010

). These energy sources are only partially predictable, because their behaviors are not fully deterministic. For example, channel fading may change the received wireless power randomly. Obstacles in the room may change the illumination too.

Figure 2.1 shows some commonly used energy sources in energy harvesting wireless communications. They have different characteristics. For example, the solar energy can only be used when or where it is sunny. The wind energy can only be used when or where it is windy. The electromagnetic energy can only be used when radio transmissions are not blocked. Hence, not all the energy sources can be used in all wireless communications systems due to size, mobility or power limitations of the wireless device. It is important to choose the appropriate source for harvesting in the designs of energy harvesting wireless communications systems. In the next section, we will discuss some of these energy sources, their characteristics, and their applications.

Figure 2.1 Some commonly used energy sources for energy harvesting wireless communications.

2.2 Types of Sources

In this section, we briefly discuss some commonly used energy sources in energy harvesting wireless communications. These sources can be mainly divided into three categories: mechanical energy, solar/light energy, and electromagnetic energy. All of them need to be converted into electricity using transducers.

2.2.1 Mechanical Energy

Mechanical energy is commonly available in our daily life. Many devices can be used to convert vibration, motion, stress, pressure or strain into electricity. Their main principle is the conversion of mechanical energy from the displacement and oscillation of a spring‐mounted mass component into electricity. Based on the randomness of the mechanical energy source, they can be categorized into three types: random vibration energy; steady flow energy; and intermittent motion energy.

The random vibration energy is often seen in built environments, such as bridges, buildings and train tracks (Roundy et al. 2003). They follow certain amplitudes and frequencies but may be random due to the random occurrence of events. The vibration energy can be extracted from these sources but the amount of energy extracted depends on the amplitude and frequency of vibration. In some cases, the presence of the energy harvesting device may also affect the vibration due to the harvester's own weight, as vibration is normally generated by the movement of a mass on a supporting frame and the harvester could add weight to the mass.

The steady flow energy comes from fluid flow, such as air or water, through pipes, or the continuous motion around a shaft, etc. Wind power is one of the most important examples of this energy. It uses the wind turbine to convert the air flow energy into electricity. Another example is the use of blood flow in vessels and breathing in human subjects to generate energy for body sensors that can monitor human body temperature or blood pressure (Mitcheson et al. 2008). The air flow and the blood flow are relatively stable so that the energy harvested from these flows is more deterministic.

The intermittent motion energy falls between vibration and flow. These energy sources come from cyclical motion in the natural environment but the energy can only be harvested during a short period of the cycle. For example, a sensor monitoring the surface of a road can harvest energy from vehicles passing over it but this energy is only available periodically. Also, energy can be harvested from human walking through shoes but only when the foot steps on the ground.

These three types of mechanical energy have different levels of randomness, leading to different levels of predictability. Based on the transduction method, the mechanical energy sources can also be categorized into three types: electromagnetic; electrostatic; and piezoelectric.

In the electromagnetic method, a magnet is used with a metal coil based on the law of induction. This method produces electricity by moving the coil through the magnetic field created by the stationary magnet (Moghe et al. 2009). When the coil moves or the distance between the magnet and the coil changes due to mechanical motion or vibration, an alternating current (AC) will be generated in the coil, which can be used to power up the wireless devices. This motion can be either controllable or uncontrollable. The advantage of this method is that no contact between coil and magnet is required and the electricity generated can be used directly. However, it is hard to integrate the electromagnetic device with the sensor circuit due to its size.

In the electrostatic method, the mechanical motion or vibration is used to change the distance between two electrodes of a capacitor against an electrical field (He et al. 2009). This will change the capacitance of a variable capacitor. The variable capacitor is made of two plates, one fixed and one moving. It needs to be initially charged. When vibration or motion separates the two plates, the vibration or motion is transformed into electricity due to the capacitance change, as the voltage across the capacitor will also change to generate a current in the circuit for use. This method allows the integration of the harvesting device into the sensor circuit.

In the piezoelectric method, a layer of piezoelectric material is used on top of the wireless device so that mechanical strain on the surface of the wireless device will be converted into electricity (Mitcheson et al. 2008). It uses a cantilever structure with a mass attached to a piezoelectric beam that has contacts on both sides of the piezoelectric material. The strain creates charge separation across the device to generate a voltage proportional to the stress applied. In some cases, the amount of energy harvested from this method is small and therefore, it may need to be combined with other methods. Also, the piezoelectric materials are breakable.

Different motion, vibration and strain sources have different power densities. For example, for a wind turbine operating at a wind speed between 2 m/s and 9 m/s, it can generate a power of about 100 mW (Ramasur and Hancke 2012). The blood flow can generate a power of 1 W, while the running shoes can generate a power of several milliwatts (Paradiso and Starner 2005; Mitcheson et al. 2008). Finger typing can generate a power of 7 mW, while lower limb movement could generate a power of 67 W (Mitcheson et al. 2008). Also, these energy sources have different applications. For example, for a wireless sensor, it is unlikely to use a wind turbine or any harvesters based on the electromagnetic method due to their bulky sizes. On the other hand, the piezoelectric method is well suited for the sensor networks due to their size but only for low‐power applications due to the limited power.

2.2.2 Solar/Light Energy

Light is perhaps one of the most commonly used sources of energy for harvesting. The photons from the light source can be converted into electricity using photovoltaic cells. The photovoltaic cell has two types of semiconductor materials and their area of contact forms a PN junction. When the photons arrive from the light source, the photovoltaic cell will release electrons to produce electricity.

For outdoor applications, solar energy is a very reliable source for self‐powered devices. It has been used in many wireless networks to replace batteries by providing an almost unlimited energy supply (Sitka et al. 2004). In most of these applications, a solar panel is used to convert the radiation from the sun into electricity. This method has been well established with relatively low cost and high efficiency over a wide range of wavelengths. Also, the energy level provided by a solar panel is very close to the nominal energy required by wireless devices. Specifically, the solar power density is around 1370 W/m2 when it arrives at the Earth and after attenuation and conversion, the available power density is still around 2 W/m2. However, one main disadvantage of solar energy is its heavy reliance on the weather, time, and the operating environment. For example, in the evenings when the sun goes down, there is hardly any solar energy to harvest. Also, for indoor applications, solar energy may not be available. In this case, it must be complemented by other energy sources. In general, solar energy is uncontrollable but can be predicted in standard conditions. In most cases, it can provide more power than any other energy sources and thus is suitable for power‐consuming energy‐harvesting communications applications.

For indoor applications, illumination from indoor lights is another source of energy. Its radiation is typically at the level of 1 W/m2 and given an efficiency of 15%, the converted electricity could be at the level of 0.15 W/m2. Typical values range between 10 W/cm2 and 100 W/cm2 (Wang et al. 2010). This is much smaller than the power density of the Sun. Indoor light is relatively controllable compared with the Sun but still varies depending on obstacles, distances, and operations.

Another type of energy that also uses the thermal effect is thermal energy (Leonov 2013). It uses the thermoelectric effect by converting the temperature difference between two metals or semiconductors of different materials into electricity. This is also called the Seebeck effect. Such temperature difference naturally occurs in human bodies or in certain machines. The amount of power converted depends on the thermoelectric properties of the materials and the temperature difference but in general is on the order of 10 W/cm2 to 1 mW/cm2 (Leonov 2013). This is suitable for wearable sensors, such as fitness bands and smart watches.

2.2.3 Electromagnetic Energy

The electromagnetic energy in this subsection mainly refers to RF energy. The advantage of RF energy over solar energy is that it can work under most conditions: indoor or outdoor, day or evening, sunny or cloudy. It can be as controllable as the light illumination but can also be as unpredictable as vibration. RF energy sources can be divided into two main categories: near field; and far field (Lu et al. 2015). The near field applications include magnetic resonance or inductive coupling. They are often used to charge devices in a wireless way over a very short distance. Due to the short distance, their efficiency can be higher than 80% but this efficiency decreases quickly with distance. This method is completely controllable and predictable. However, for wireless communications systems, this short distance may not be realistic. Hence, energy harvesting wireless communications often use the far field method that can harvest energy over a distance of more than 10 m.

The sources of RF energy in the far field method can be from the ambient environment, such as radiations from the cellular base station, TV transmitter or WiFi router. It can also be from dedicated power transmitters. One unique advantage of RF energy is that most wireless systems are implemented using radio waves too and hence, information delivery can be combined with energy transfer in the same system and sometimes by the same signal.

The level of power from a global system for mobile communications (GSM) base station is around −40 dBm/cm2. Studies show that other ambient sources, such as TV, third generation (3G) and WiFi produce even weaker power. For example, a 3G base station generates a power density of around −50 dBm/cm2, while WiFi signals provide a power density of around −70 dBm/cm2 (Pinuela et al. 2013). Hence, although there are many different ambient RF energy sources, in general their power densities are very low, as their power densities decay quickly with distance. Consequently, these ambient RF sources can only be used for low‐power applications, such as radio frequency identification (RFID) or sensor networks, or the wireless device must be very close to the sources. To harvest enough energy for more power‐consuming wireless operations, either a large antenna or a wide‐band antenna need to be used. Alternatively, dedicated sources of RF energy are required, as in wireless powered communications systems, at additional cost.

There are other types of energy available for harvesting. For example, the pyroelectric effect of materials can be used to generate electricity. Biomedical substances can be used to generate biochemical energy. Acoustic waves can be converted into electricity using transducers or resonators too. Alternatively, all the above energy sources can be combined. Table 2.1 gives an overview of the amount of energy available from different sources.

Table 2.1 Typical amount of energy from different sources.

Source

Typical amount of energy

Solar

100 mW/cm

2

(sunny)

Indoor light

0.01

0.1 mW/cm

2

Wind

0.38 mW/cm

3

(at 5 m/s)

Piezoelectric

0.2

0.4 mW/cm

3

Electrostatic

0.05

0.1 mW/cm

3

Ambient RF

0.2 nW/cm

2

1

W/cm

2

2.3 Predictive Models of Sources

The amount of energy from most energy sources varies with time. This time‐variance leads to uncertainty in the energy supply for wireless communications. Thus, it is very useful to have accurate models that describe this energy uncertainty, because wireless communications systems can use these models to make critical decisions on the usage of energy.

In this section, we will discuss some important modeling studies on the amount of energy provided by various sources. The data used to derive these models were collected by using a certain measuring equipment or energy harvester. Thus, there is a conversion loss from the available energy at the source as the input of the equipment to the measured or harvested energy as the output of the equipment. Nevertheless, since the same equipment or energy harvester is used to collect all data, this conversion loss can be considered as constant so that the behaviors of the data before collection and after collection are approximately the same to justify the usefulness of the derived models. In the next chapter, we will discuss models that describe the conversion loss or the efficiency of the energy harvester. Using the energy source model here and the energy harvester model in the next chapter, one can predict how much energy is available for wireless communications. We will also discuss models of the harvested power directly in the next chapter. In the following, we will focus on the solar energy models and the ambient RF energy models, as they are the two most widely used energy sources in wireless communications systems.

2.3.1 Solar Energy Modeling

As discussed before, solar energy is not controllable but due to its clear diurnal and seasonal patterns, it is predictable. However, its prediction is highly dependent on the weather conditions.

The simplest model for solar energy prediction is the exponentially weighted moving average (EWMA) model (Cox 1961; Kansal et al. 2007). It divides the data at different time slots on different days into a matrix, where the columns of the matrix could represent different time slots on the same day while the rows of the matrix could represent different days. It uses a weighting factor of to predict the solar energy in the next time slot by linearly combining the current measurement and the previously predicted solar energy. The weighting factor decreases with time to lay a higher emphasis on the measurements taken at a time closer to the time to be predicted. Mathematically, the EWMA predictor is given by

(2.1)

where is the predicted value in the next time slot , is the measurement in the current time slot , and is the predicted value in the current time slot . The exponential weighting can be seen by replacing with to give

(2.2)

where the predicted value at time slot has a smaller weighting factor of since . Using 2.1 in solar energy prediction, one has

(2.3)

where represents the day, represents the time slot on that day, is the predicted value at time on the th day, is the measured value at time slot on the th day, and is the predicted value at time slot on the th day. If a matrix is used, is the element on the th row and th column of the measurement matrix. From 2.3, the prediction for each time slot is calculated by taking into account the predicted and measured values at the same time slot on the previous day.

The EWMA model works well when the weather condition is stable over a few days or does not change at all. However, if the weather does change, its accuracy will decrease. For example, if the weather keeps switching between sunny and cloudy on different days, using the predicted and measured values on the current day will not help the prediction for the next day much. In this case, the weather conditioned moving average (WCMA) model can be used (Piorno et al. 2009). The WCMA model again divides data into a matrix with rows representing days and columns representing time slots. However, unlike EWMA, it uses the average of the measured values on a few previous days, not just one previous day. Specifically, one has

(2.4)

where is the value predicted for the next time slot on the th day, is the measured value at the current time slot on the th day, is the number of previous days used, and is the measured value at time slot on the th day, . This model can improve the accuracy over the EWMA model by using an average of the values at the same time slots on previous days instead of one single predicted value on the previous day. However, if there is a cloudy day followed by many sunny days or vice versa, this method will cause errors. To reduce the error, the measurements in the previous time slots on the same day can also be used, to replace a single measurement at the current time slot on the same day assuming that the weather conditions are at least stable for the whole day. In this case, the WCMA model can be modified as

(2.5)

where

(2.6)

is the number of time slots in the past on the same day, is a vector with the th element being , is a vector with the th element being , is the dot product of two vectors, and . One sees that there is an additional weighting factor to consider the variance over different time slots.

Another simple predictor uses a 2‐D linear filter. In this case, the predicted value is calculated as

(2.7)

where the parameters of , and