Table of Contents

Cover

Title Page

About the Authors

Preface

Acknowledgments

Chapter 1: Characterization of Wireless Transmitter Distortions

1.1 Introduction
1.2 Impact of Distortions on Transmitter Performances
1.3 Output Power versus Input Power Characteristic
1.4 AM/AM and AM/PM Characteristics
1.5 1 dB Compression Point
1.6 Third and Fifth Order Intercept Points
1.7 Carrier to Inter-Modulation Distortion Ratio
1.8 Adjacent Channel Leakage Ratio
1.9 Error Vector Magnitude
References

Chapter 2: Dynamic Nonlinear Systems

2.1 Classification of Nonlinear Systems
2.2 Memory in Microwave Power Amplification Systems
2.3 Baseband and Low-Pass Equivalent Signals
2.4 Origins and Types of Memory Effects in Power Amplification Systems
2.5 Volterra Series Models
References

Chapter 3: Model Performance Evaluation

3.1 Introduction
3.2 Behavioral Modeling versus Digital Predistortion
3.3 Time Domain Metrics
3.4 Frequency Domain Metrics
3.5 Static Nonlinearity Cancelation Techniques
3.6 Discussion and Conclusion
References

Chapter 4: Quasi-Memoryless Behavioral Models

4.1 Introduction
4.2 Modeling and Simulation of Memoryless/Quasi-Memoryless Nonlinear Systems
4.3 Bandpass to Baseband Equivalent Transformation
4.4 Look-Up Table Models
4.5 Generic Nonlinear Amplifier Behavioral Model
4.6 Empirical Analytical Based Models
4.7 Power Series Models
References

Chapter 5: Memory Polynomial Based Models

5.1 Introduction
5.2 Generic Memory Polynomial Model Formulation
5.3 Memory Polynomial Model
5.4 Variants of the Memory Polynomial Model
5.5 Envelope Memory Polynomial Model
5.6 Generalized Memory Polynomial Model
5.7 Hybrid Memory Polynomial Model
5.8 Dynamic Deviation Reduction Volterra Model
5.9 Comparison and Discussion
References

Chapter 6: Box-Oriented Models

6.1 Introduction
6.2 Hammerstein and Wiener Models
6.3 Augmented Hammerstein and Weiner Models
6.4 Three-Box Wiener–Hammerstein Models
6.5 Two-Box Polynomial Models
6.6 Three-Box Polynomial Models
6.7 Polynomial Based Model with I/Q and DC Impairments
References

Chapter 7: Neural Network Based Models

7.1 Introduction
7.2 Basics of Neural Networks
7.3 Neural Networks Architecture for Modeling of Complex Static Systems
7.4 Neural Networks Architecture for Modeling of Complex Dynamic Systems
7.5 Training Algorithms
7.6 Conclusion
References

Chapter 8: Characterization and Identification Techniques

8.1 Introduction
8.2 Test Signals for Power Amplifier and Transmitter Characterization
8.3 Data De-Embedding in Modulated Signal Based Characterization
8.4 Identification Techniques
8.5 Robustness of System Identification Algorithms
8.6 Conclusions
References

Chapter 9: Baseband Digital Predistortion

9.1 The Predistortion Concept
9.2 Adaptive Digital Predistortion
9.3 The Predistorter's Power Range in Indirect Learning Architectures
9.4 Small Signal Gain Normalization
9.5 Digital Predistortion Implementations
9.6 The Bandwidth and Power Scalable Digital Predistortion Technique
9.7 Summary
References

Chapter 10: Advanced Modeling and Digital Predistortion

10.1 Joint Quadrature Impairment and Nonlinear Distortion Compensation Using Multi-Input DPD
10.2 Modeling and Linearization of Nonlinear MIMO Systems
10.3 Modeling and Linearization of Dual-Band Transmitters
10.4 Application of MIMO and Dual-Band Models in Digital Predistortion
References

Index

End User License Agreement

List of Illustrations

Chapter 1: Characterization of Wireless Transmitter Distortions

Figure 1.1 Simplified block diagram of a typical wireless transmitter
Figure 1.2 Gain and power efficiency characteristics of a power amplifier prototype
Figure 1.3 Frequency domain output of a nonlinear transmitter driven by a two-tone signal
Figure 1.4 Output spectrum of a nonlinear transmitter driven by a multi-carrier WCDMA signal
Figure 1.5 Gain and efficiency considerations in brute force linear amplification
Figure 1.6 Gain and efficiency considerations in linearized power amplifiers
Figure 1.7 Sample output power versus input power characteristic
Figure 1.8 Sample AM/AM characteristic of a power amplifier
Figure 1.9 Sample AM/PM characteristic of a power amplifier
Figure 1.10 Graphical definition of the 1 dB compression point from P_out vs. P_in characteristic
Figure 1.11 Graphical definition of the 1 dB compression point from the AM/AM characteristic
Figure 1.12 Output power characteristics at the fundamental and third order inter-modulation frequencies
Figure 1.13 Graphical definition of the third order intercept point
Figure 1.14 Graphical definition of the carrier to inter-modulation distortion ratio
Figure 1.15 Graphical definition of the adjacent channel leakage ratio
Figure 1.16 Effects of phase and amplitude distortions on the QPSK constellation. (a) Effects of phase distortions. (b) Effects of amplitude distortions. (c) Effects of phase and amplitude distortions
Figure 1.17 Graphical definition of the error vector

Chapter 2: Dynamic Nonlinear Systems

Figure 2.1 Example of an AM/AM and AM/PM curve for a memoryless system
Figure 2.2 Example of an AM/AM and AM/PM curve for a system with memory
Figure 2.3 Pass-band and baseband signals
Figure 2.4 Block diagram of a common source MESFET amplifier showing the definition of the different impedances
Figure 2.5 Modeling of nonlinear electrical memory effects in a cascade of two nonlinear systems
Figure 2.6 Modeling of thermal memory effects in a power transistor. (a) Different temperatures definitions. (b) Circuit modeling of the temperature variation
Figure 2.7 Simplified transistor thermal modeling
Figure 2.8 Junction temperature variation for a step input. (a) Input signal variation versus time. (b) Corresponding junction temperature variation versus time
Figure 2.9 Junction temperature variation for a pulsed input. (a) Input signal variation versus time. (b) Corresponding junction temperature variation versus time
Figure 2.10 Complex gain variation as a function of the junction temperature. (a) Gain in dB. (b) Phase shift in degrees
Figure 2.11 Generic block diagram for behavioral model performance assessment

Chapter 3: Model Performance Evaluation

Figure 3.1 Generic block diagram for behavioral model performance assessment
Figure 3.2 Generic block diagram of a digital predistortion system
Figure 3.3 Generic block diagram for digital predistorter performance assessment (a) considering the DPD and (b) considering the linearized DUT
Figure 3.4 Common representation of model identification variables in time domain
Figure 3.5 Block diagram of the static nonlinearity pre-compensation technique
Figure 3.6 Frequency domain performance of behavioral models (case of a memoryless DUT) (a) before memoryless pre-compensation and (b) after memoryless pre-compensation
Figure 3.7 Frequency domain performance of behavioral models (case of a DUT with memory) (a) before memoryless pre-compensation and (b) after memoryless pre-compensation
Figure 3.8 Block diagram of the static nonlinearity post-compensation technique
Figure 3.9 Frequency domain performance of behavioral models (case of a memoryless DUT) (a) before memoryless post-compensation and (b) after memoryless post-compensation
Figure 3.10 Frequency domain performance of behavioral models (case of a DUT with memory) (a) before memoryless post-compensation and (b) after memoryless post-compensation
Figure 3.11 Graphical illustration of the MEI integration bandwidths

Chapter 4: Quasi-Memoryless Behavioral Models

Figure 4.1 Illustration of the amplitude and phase transfer characteristics of an envelope nonlinearity of a TWTA
Figure 4.2 Block diagram of the behavioral model for AM/AM and AM/PM envelope nonlinearity; both $c04-math-0015$ and $c04-math-0016$ are functions of time. (a) Symbolic model at carrier frequency. (b) Explicit model at complex envelope level
Figure 4.3 Quadrature model of envelope quasi-memoryless nonlinearity. (a) Symbolic model at carrier frequency. (b) Explicit model at complex envelope level
Figure 4.4 LUT for quasi-memoryless system modeling
Figure 4.5 Cartesian implementation of the LUT based model for quasi-memoryless nonlinearity
Figure 4.6 Non-uniform LUT architecture
Figure 4.7 Companding functions for LUTs
Figure 4.8 Compression coefficient k as function of $c04-math-0116$ and $c04-math-0117$
Figure 4.9 Saleh model with parameters: $c04-math-0137$ , $c04-math-0138$ , $c04-math-0139$ , and $c04-math-0140$
Figure 4.10 Complex gain profile for Saleh model with parameters: $c04-math-0141$ , $c04-math-0142$ , $c04-math-0143$ , and $c04-math-0144$
Figure 4.11 Gain and phase profiles for Ghorbani model for GaAsFET SSPA
Figure 4.12 Phase shift transfer function characteristic of the Berman and Mahle model
Figure 4.13 AM/AM and AM/PM characteristics for TWTA modeled using the Thomas–Weidner–Durrani model
Figure 4.14 Limiter characteristics
Figure 4.15 ARCTAN model characteristics
Figure 4.16 Rapp model characteristics for different smoothing factor $c04-math-0224$
Figure 4.17 AM/AM and AM/PM characteristics of the White model

Chapter 5: Memory Polynomial Based Models

Figure 5.1 Trends in single-box memory polynomial models development
Figure 5.2 Block diagram of the memory polynomial model
Figure 5.3 Block diagram of the orthogonal memory polynomial model
Figure 5.4 Block diagram of the sparse-delay memory polynomial model
Figure 5.5 Block diagram of the non-uniform memory polynomial model
Figure 5.6 Structure of memory polynomial model variants
Figure 5.7 Block diagram of the envelope memory polynomial model
Figure 5.8 Complex gain based representation of the envelope memory polynomial model
Figure 5.9 Block diagram of the generalized memory polynomial model. (a) Generalized memory polynomial model. (b) Time-aligned terms memory polynomial (c) Lagging cross-terms memory polynomial (d) Leading cross-terms memory polynomial
Figure 5.10 Measured spectra at the output of a Doherty amplifier linearized using the memory polynomial and the generalized memory polynomial DPDs
Figure 5.11 Block diagram of the hybrid memory polynomial model
Figure 5.12 Predicted spectra at the output of a power amplifier using the memory polynomial, the envelope memory polynomial, and the hybrid memory polynomial models. (a) Full spectrum and (b) zoomed version
Figure 5.13 Comparison between memory polynomial based models. (a) Weakly nonlinear memory effects, (b) mildly nonlinear memory effects, and (c) strongly nonlinear memory effects

Chapter 6: Box-Oriented Models

Figure 6.1 Block diagram of the Wiener model
Figure 6.2 Block diagram of the Hammerstein model
Figure 6.3 Augmented Wiener model
Figure 6.4 Augmented Hammerstein model
Figure 6.5 Wiener–Hammerstein model
Figure 6.6 Hammerstein–Wiener model
Figure 6.7 Feedforward Hammerstein model
Figure 6.8 Twin nonlinear twin box models. (a) Forward TNTB model, (b) reverse TNTB model, and (c) parallel TNTB model
Figure 6.9 PLUME model
Figure 6.10 Three layered biased memory polynomial model
Figure 6.11 Parallel Hammerstein based model for the alleviation of PA nonlinearity and I/Q imbalance
Figure 6.12 Two-box model with I/Q and DC impairments

Chapter 7: Neural Network Based Models

Figure 7.1 Block diagram of a typical L-layer feedforward neural network
Figure 7.2 Most common activation functions used in artificial neural network
Figure 7.3 Classification of neural networks structures for power amplifiers modeling and predistortion applications
Figure 7.4 Single-input single-output feedforward neural network
Figure 7.5 Block diagram of dual-input dual-output feedforward neural network applied on the polar components
Figure 7.6 Dual-input dual-output coupled Cartesian based neural network
Figure 7.7 Block diagram of a two-layer complex time-delay recurrent neural network (CTDRNN)
Figure 7.8 Block diagram of complex time delay neural network (CTDNN)
Figure 7.9 Block diagram of a three-layer real valued time delay recurrent neural network (RVTDRNN)
Figure 7.10 Block diagram of a three-layer real valued focused time-delay neural network
Figure 7.11 Predicted output voltage magnitude of a Doherty power amplifier prototype. (a) RVRNN model and (b) RVFTDNN model [12].
Figure 7.12 Predicted output voltage phase of a Doherty power amplifier prototype. (a) RVRNN model and (b) RVFTDNN model [12].
Figure 7.13 Comparison of training algorithms performance for RVFTDNN based digital predistorter [12].

Chapter 8: Characterization and Identification Techniques

Figure 8.1 Flow chart of behavioral modeling and digital predistortion processes
Figure 8.2 Generic block diagram of experimental setup for power amplifiers characterization
Figure 8.3 Block diagram of experimental setup for nonlinear behavior characterization through complex baseband waveforms measurements. (a) PA only characterization and (b) analog front-end and PA characterization
Figure 8.4 Measured AM/AM and AM/PM characteristics of a power amplifier prototype for various excitation signals. (a) AM/AM characteristics and (b) AM/PM characteristics [14].
Figure 8.5 Measured AM/AM and AM/PM characteristics of a power amplifier prototype for WCDMA signals with various PAPR. (a) AM/AM characteristics. (b) AM/PM characteristics
Figure 8.6 Measured AM/AM and AM/PM characteristics of a power amplifier prototype for WCDMA signal with various average power levels. (a) AM/AM characteristics and (b) AM/PM characteristics
Figure 8.7 Measured AM/AM characteristics of a power amplifier prototype for LTE signals with various bandwidths. (a) One-carrier (20 MHz bandwidth) and (b) three-carrier (60 MHz bandwidth)
Figure 8.8 PA characterization results before delay compensation. (a) Time domain waveforms and (b) AM/AM characteristic
Figure 8.9 PA characterization results after coarse delay compensation. (a) Time domain waveforms and (b) AM/AM characteristic
Figure 8.10 PA characterization results after fine delay compensation. (a) Time domain waveforms and (b) AM/AM characteristic
Figure 8.11 Block diagram of the system identification concept in adaptive filtering
Figure 8.12 Block diagram of the system reverse modeling concept in adaptive filtering

Chapter 9: Baseband Digital Predistortion

Figure 9.1 Simplified block diagram of predistortion system and corresponding gain characteristics of each block
Figure 9.2 Power transfer characteristics involved in a predistortion system
Figure 9.3 Power transfer characteristics in predistortion systems
Figure 9.4 Closed loop adaptive digital predistortion system
Figure 9.5 Open loop adaptive digital predistortion system
Figure 9.6 Open loop adaptive digital predistortion system implementations (a) direct learning architecture and (b) indirect learning architecture
Figure 9.7 Power transfer characteristics of a PA and its predistorter
Figure 9.8 Characteristic curves of a power amplifier and its predistorter, (a) power transfer characteristic of the power amplifier before and after linearization and (b) AM/AM characteristic of the predistorter
Figure 9.9 Impact of the normalization gain on the DPD and linearized DUT characteristics, (a) DPD's AM/AM characteristics, (b) DPD's power transfer characteristics
Figure 9.10 Effects of gain normalization on the average power variation through a digital predistorter
Figure 9.11 Effects of the average power variation through digital predistorter on the ACPR of a linearized power amplifier prototype [7].
Figure 9.12 Effects of the DPD normalization gain on the output spectra of a linearized power amplifier prototype [7].
Figure 9.13 Baseband digital predistortion system implementation
Figure 9.14 Measurement based digital predistortion test bed
Figure 9.15 Hybrid RF digital predistortion system
Figure 9.16 Digital predistortion system update. (a) Conventional approach and (b) bandwidth and power scalable approach (c) LDUT power transfer characteristics

Chapter 10: Advanced Modeling and Digital Predistortion

Figure 10.1 Modeling of the I/Q modulator's imbalance
Figure 10.2 Dual-input nonlinear model for the joint effects of quadrature impairments and PA nonlinearities
Figure 10.3 Block diagram of a linear and nonlinear distortion compensation composed of a cascade of a Hammerstein PA predistorter and a quadrature imbalance compensator [11]
Figure 10.4 Concept of transforming a single-branch serial structure (cascade of a Hammerstein PA predistorter and a quadrature imbalance compensator) to a parallel dual-branch parallel Hammerstein structure: (a) the original structure of the digital predistorter and quadrature imbalance compensator in one branch of the PH model; (b) the modified structure after merging the frequency dependent LTI parts; and (c) the final structure after splitting the static nonlinearity between the conjugate and non-conjugate branches
Figure 10.5 Detailed block diagram of a parallel dual-branch parallel Hammerstein structure
Figure 10.6 Detailed block diagram of the dual-conjugate-input memory polynomial structure
Figure 10.7 Linear and nonlinear crosstalks in MIMO transmitters
Figure 10.8 Linear crosstalk model in MIMO configuration
Figure 10.9 Conventional digital predistortion in the presence of nonlinear crosstalk in a MIMO transmitter
Figure 10.10 Effect of power amplifier nonlinearity on concurrent dual-band transmission
Figure 10.11 Performance of conventional single-input single-output linearization in the presence of nonlinear crosstalk in MIMO systems [21].
Figure 10.12 Performance of MIMO DPD in the presence of nonlinear crosstalk in MIMO systems [21]. ©2009 IEEE. Reprinted, with permission, from S. A. Bassam et al.,
Figure 10.13 Performance of 2-D-DPD linearization in the presence of concurrent dual-band transmission (a) band I with a 10 MHz OFDM signal (©2009 IEEE. Reprinted, with permission, from S. A. Bassam et al., “2-D digital predistortion (2-D-DPD) architecture for concurrent dual-band transmitters,” IEEE Transactions on Microwave Theory and Techniques, Oct. 2011) and (b) band II with a 5 MHz WCDMA signal
Figure 10.14 Performance of 3-D phase-aligned Volterra DPD linearization in the presence of concurrent tri-band-band transmission and comparison with the 3-D memory polynomial DPD (a) band I; (b) band II; and (c) band III [44].

List of Tables

Chapter 1: Characterization of Wireless Transmitter Distortions

Table 1.1 Frequency components at the output of a nonlinear transmitter for a two-tone input signal

Chapter 4: Quasi-Memoryless Behavioral Models

Table 4.1 First five Gegenbauer polynomial basis functions
Table 4.2 First nine basis functions $c04-math-0314$

Chapter 5: Memory Polynomial Based Models

Table 5.1 Comparison between memory polynomial models' complexity

Chapter 8: Characterization and Identification Techniques

Table 8.1 Performance comparison of LS, RLS, and LMS algorithms

Chapter 10: Advanced Modeling and Digital Predistortion

Table 10.1 EVM and ACPR performance of MIMO DPD compared to conventional single-band linearization in the presence of nonlinear crosstalk in MIMO systems [21]
Table 10.2 NMSE and ACPR performance of 2-D-DPD compared to conventional single-band linearization in the case of concurrent dual-band transmission [37]
Table 10.3 ACPR performance of 2-D-DPD compared to conventional single-band linearization in the case of concurrent dual-band transmission [44]

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

For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com.

The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988.

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 the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books.

Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. It is sold on the understanding that the publisher is not engaged in rendering professional services and neither the publisher nor the author shall be liable for damages arising here from. If professional advice or other expert assistance is required, the services of a competent professional should be sought.

Behavioral modelling and predistortion of wideband wireless transmitters / Fadhel Ghannouchi, Oualid Hammi, Mohamed Helaoui.

ISBN 978-1-118-40627-4 (cloth : alk. paper) 1. Wireless communication systems–Mathematical models. 2. Broadband communication systems–Mathematical models. 3. Signal theory (Telecommunication)–Mathematics. 4. Telecommunication–Transmitters and transmission–Mathematics. 5. Nonlinear systems–Mathematical models. 6. Electric distortion–Mathematical models. I. Title.

About the Authors

Dr Fadhel M. Ghannouchi was born in Gabes, Tunisia in 1958. He is currently Professor and iCORE/CRC Chair at the Electrical and Computer Engineering Department of The Schulich School of Engineering of the University of Calgary and founding Director of Intelligent RF Radio Laboratory (www.iradio.ucalgary.ca). His research interests are in the areas of microwave instrumentation and measurements, nonlinear modeling of microwave devices and communications systems, design of power and spectrum efficient microwave amplification systems, and design of intelligent RF transceivers for wireless and satellite communications. His research activities have led to over 600 publications and 15 US patents (three pending) and four books. He is the co-founder of three of the University's spin-off companies. As an educator and mentor, he has supervised and graduated more than 80 Masters and PhD students and has supervised more than 30 post-doctoral fellows. Dr Ghannouchi is an IEEE Fellow, IET Fellow, Fellow of Engineering Institute of Canada (EIC), Fellow of the Canadian Academy of Engineering (CAE), and Fellow of the Royal Society of Canada (RSC). Dr Ghannouchi has also been an IEEE Distinguish Microwave Lecturer (2009–2012).

Dr Oualid Hammi was born in Tunis, Tunisia in 1978. He received a B.Eng. degree from the École Nationale d'Ingénieurs de Tunis, Tunis, Tunisia, in 2001, an MSc degree from the École Polytechnique de Montréal, Montréal, QC, Canada, in 2004, and his PhD from the University of Calgary, Calgary, AB, Canada, in 2008, all in electrical engineering. He is currently an Associate Professor with the Department of Electrical Engineering, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia. He is also an adjunct researcher with the Intelligent RF Radio Laboratory (iRadio Lab), Schulich School of Engineering, University of Calgary. He has authored and co-authored over 80 publications and 7 US patents (six pending), and is a regular reviewer for several IEEE transactions and other journals. His research interests include the characterization, behavioral modeling, and linearization of radiofrequency power amplifiers and transmitters, and the design of energy efficient linear transmitters for wireless communication systems.

Mohamed Helaoui was born in Tunis, Tunisia in 1979. He received his MSc degree in communications and information technology from the École Supérieure des Communications de Tunis, Tunisia, in 2003 and his PhD in electrical engineering from the University of Calgary in 2008. He is currently an Associate Professor at the Department of Electrical and Computer Engineering at the University of Calgary. His current research interests include digital signal processing, power efficiency enhancement for wireless transmitters, switching mode power amplifiers, six-port receivers, and advanced transceiver design for software defined radio and millimeter-wave applications. His research activities have led to over 100 publications and eight patents (three pending).

Preface

Wireless systems are offering a wide variety of services to an ever increasing number of users. Undeniably, this connectivity has contributed to enhancing the quality of life. Though, the proliferation of wireless handheld devices and base stations led to an alarming downside due to their environmental impact. In fact, the carbon footprint of the wireless communication infrastructure is reaching unprecedented levels. This stimulated a global awareness about the need to reduce base stations energy consumption. In order to make communication systems more eco-friendly and “greener”, significant research work is being carried out at various aspects of base station design. This includes, among other things, scaling of energy needs depending on the traffic and network load, improving the ratio of quality of service to radiofrequency power, and increasing the overall efficiency of the base station. A closer look at base stations power consumption reflects that their overall efficiency can be significantly improved by increasing that of the radio frequency front end and especially the power amplifier. This would not only make communication systems greener but also reduce their deployment and running costs in terms of capital expenditure (CAPEX) and operational expenditure (OPEX), and result in substantial financial benefits.

Technically, building power amplifiers with peak power efficiencies as high as 80% has become feasible thanks to the development of new transistor technologies and new classes of operation such as switching mode. However, getting such high efficiencies from power amplifiers handling modern wireless communication systems is a tricky challenge. In fact, and due to the nature of the highly varying envelop signals being transmitted, base station power amplification systems have to be highly linear and meet the spectrum emission masks set by standardization and regulatory authorities. This requires the use of linearization techniques, which virtually make the power amplifier linear over its entire power range, thus allowing operation with less power back-off, and hence resulting in higher efficiencies compared to what could have been obtained from the same amplifier if no linearization was adopted. In this context, digital predistortion has received tremendous attention from the industrial and academic communities and incontestably appears to be the preferred technology for base station power amplifier linearization.

Conceptually, behavioral modeling and digital predistortion are intimately related. They are often referred to as forward and reverse modeling, respectively. This book focuses on the behavioral modeling and digital predistortion of wideband power amplifiers and transmitters. It compiles a wide range of topics related to this theme. The book is organized in 10 chapters, which can be organized into three parts. Chapters 1–3 set the ground for the remainder of the book by introducing the key parameters used to model and characterize the nonlinear behavior of wireless transmitters in Chapter 1, classifying and discussing the theory of dynamic nonlinear systems in Chapter 2, and providing a review of model performance evaluations metrics in Chapter 3. The second part of the book, Chapters 4–7, is a thorough review of behavioral models and predistortion functions that encompasses quasi-memoryless models in Chapter 4, memory polynomial based models in Chapter 5, box-oriented models in Chapter 6, and neural networks based models in Chapter 7. These models are introduced and their specificities discussed. The last part of the book, Chapters 8–10, is application oriented and provides comprehensive and insightful information about the use, in an experimental environment, of the models described earlier in the book. Chapter 8 covers the acquisition of the device-under-test (DUT) input and output data and its processing prior to the model identification. Chapter 9 is devoted to baseband digital predistortion and its practical aspects. Chapter 10 concludes the book by exposing recent trends in behavioral modeling and digital predistortion such as joint quadrature impairment compensation and digital predistortion, as well as the predistortion of dual-band and multi-input multi-output (MIMO) transmitters.

The book chapters are complemented with a software tool available through the Wiley website (www.wiley.com/go/Ghannouchi/Behavioral) that implements several of the topics discussed in the book and can be used to demonstrate these topics in a more tangible way.

Acknowledgments

We would like to gratefully acknowledge the help and support received from friends, colleagues, support staff, and students, both past and present at iRadio Laboratory, University of Calgary, Calgary; Poly-GRAMES Research Center, Ecole Polytechnique, Montreal; and King Fahd University of Petroleum and Minerals, KSA. We are grateful to our great students and researchers; this book could not have been completed without their fruitful research. In addition, we would like to thank C. Heys and A. Congreve for their help in proofreading and formatting the manuscript, and Ivana d'Adamo for her administrative support. The authors would also like to thank the IEEE for their acceptance and courtesy in allowing them to reproduce several figures and illustrations published elsewhere in their journals and/or conferences.

Dr O. Hammi acknowledges the support provided by the Deanship of Scientific Research at King Fahd University of Petroleum and Minerals (KFUPM) under research grant IN121039.

Dr F. M. Ghannouchi and M. Helaoui acknowledge the main sponsors and financial supporters of the Intelligent Radio technology Laboratory (iRadio Lab), Alberta Innovates Technology Futures (AITF), Alberta, Canada, the Canada Research Chairs (CRC) program, and Natural Sciences and Engineering Research Council of Canada (NSERC).

Finally, we would like to profoundly thank our respective spouses; Ilhem, Asma, and Imen; and our kids for their understanding and patience throughout the many evenings and weekends taken to prepare this book, as well as our parents for their encouragement and valuable support in our early professional years as graduate students and young researchers.