UNDERWATER ACOUSTIC SIGNAL PROCESSING AND ITS APPLICATIONS By KEXIN ZHAO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2014 ⃝c 2014 Kexin Zhao 2 I dedicate this work to my Mother, Ling Cao, and my fiance´e, Yixue Zhang, without them this would have been impossible for me. 3 ACKNOWLEDGMENTS First and foremost, I would like to express my deepest appreciation to my advisor, Dr. Jian Li of the Electrical and Computer Engineering Department, who has led me throughout my PhD training with her patience, great knowledge, and continuous encouragement. It has been an exceptional experience to participate in the underwater acoustic communication and the active sonar projects. Without her guidance and persistent help this dissertation would not have been possible. I acknowledge my committee members at University of Florida: Dr. Mingzhou Ding, Dr. Tan Wong, and Dr. Jenshan Lin. I am truly thankful for their time and efforts that they spent on my dissertation. I am truly grateful to have a nice group of fellow students at Dr. Li’s lab. I highly appreciate all the help that I received from my lab members during my time at University of Florida. Research would have been much less colorful without them. Last but not least, I am deeply indebted to my family, especially my mother. They have provided me with immense understanding and support all these years. This dissertation is dedicated to them. 4 TABLE OF CONTENTS page ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 CHAPTER 1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 1.1 Challenges of UAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.2 Challenges of Active Sonar Systems . . . . . . . . . . . . . . . . . . . . . 14 1.3 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2 ENHANCED MOBILE MULTI-INPUT MULTI-OUTPUT UAC . . . . . . . . . . . 22 2.1 System Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2.2 Double-Selective Channel with Doppler Scaling Effects . . . . . . . . . . 26 2.2.1 Channel Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 2.2.2 Temporal Resampling . . . . . . . . . . . . . . . . . . . . . . . . . 28 2.2.3 Resampling Factor Estimation . . . . . . . . . . . . . . . . . . . . . 28 2.3 Channel Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.3.1 Training-Directed Mode . . . . . . . . . . . . . . . . . . . . . . . . 29 2.3.2 Decision-Directed Mode . . . . . . . . . . . . . . . . . . . . . . . . 31 2.3.3 Channel Estimation Algorithm: GoSLIM . . . . . . . . . . . . . . . 32 2.3.4 Channel Estimation Algorithm: GoSLIM-V . . . . . . . . . . . . . . 35 2.3.5 Complexity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 36 2.4 Symbol Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 2.4.1 Problem formulation . . . . . . . . . . . . . . . . . . . . . . . . . . 38 2.4.2 Phase Compensation . . . . . . . . . . . . . . . . . . . . . . . . . 39 2.4.3 LMMSE Based Soft-Input Soft-Output Equalizer . . . . . . . . . . . 40 2.4.3.1 A priori LLR pre-processor . . . . . . . . . . . . . . . . . 40 2.4.3.2 LMMSE filtering . . . . . . . . . . . . . . . . . . . . . . . 41 2.4.3.3 A posteriori LLR generator . . . . . . . . . . . . . . . . . 43 2.4.4 Low-Complexity Approximate LMMSE Filtering . . . . . . . . . . . 44 2.5 Numerical and Experimental Results . . . . . . . . . . . . . . . . . . . . . 45 2.5.1 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 2.5.2 MACE10 In-Water Experimentation Results . . . . . . . . . . . . . 47 2.5.2.1 Experiment specifics . . . . . . . . . . . . . . . . . . . . . 47 2.5.2.2 Performance evaluation . . . . . . . . . . . . . . . . . . . 49 5 3 ENHANCED MULTISTATIC ACTIVE SONAR SIGNAL PROCESSING . . . . . 64 3.1 System Description and Problem Formulation . . . . . . . . . . . . . . . . 66 3.2 Proposed Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 3.2.1 Range-Doppler Imaging . . . . . . . . . . . . . . . . . . . . . . . . 68 3.2.1.1 Imaging problem formulation . . . . . . . . . . . . . . . . 68 3.2.1.2 Receiver filter for range-Doppler imaging . . . . . . . . . 70 3.2.2 Generalized K-Means Clustering (GKC) Association Method . . . . 74 3.2.3 EXIP-WLS Method for Target Position Estimation . . . . . . . . . . 77 3.2.4 EXIP-WLS Method for Target Velocity Estimation . . . . . . . . . . 80 3.3 Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 3.3.1 Range-Doppler Imaging Results . . . . . . . . . . . . . . . . . . . 83 3.3.2 Target Parameter Estimation Results . . . . . . . . . . . . . . . . . 83 4 WIDEBAND SOURCE LOCALIZATION USING SLIM . . . . . . . . . . . . . . 91 4.1 Data Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 4.2 The Wideband SLIM Algorithms . . . . . . . . . . . . . . . . . . . . . . . 96 4.2.1 WB-SLIM-0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 4.2.2 WB-SLIM-1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 4.2.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 4.3 RELAX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 4.4 Numerical Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 5 CONCLUSIONS AND FUTURE WORK . . . . . . . . . . . . . . . . . . . . . . 112 APPENDIX A THE DERIVATION OF THE DOPPLER SCALING FACTOR . . . . . . . . . . . 115 B THE DERIVATION OF THE WEIGHT MATRIX . . . . . . . . . . . . . . . . . . 116 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 BIOGRAPHICAL SKETCH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 6 LIST OF TABLES Table page 2-1 Coded BER results obtained by GoSLIM and GoSLIM-V . . . . . . . . . . . . . 53 2-2 Complexity comparison (in s) between GoSLIM and GoSLIM-V . . . . . . . . . 53 2-3 A summary of the performance of the three detection schemes . . . . . . . . . 54 2-4 The average coded BER obtained by Exact LMMSE Turbo and Approximate LMMSE Turbo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 3-1 The noise power and the norm of the target reflection coefficients . . . . . . . . 85 3-2 System parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 3-3 RMSE of Parameter Estimates Using ULS and EXIP-WLS . . . . . . . . . . . . 85 4-1 WB-SLIM algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 4-2 The RELAX algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 7 LIST OF FIGURES Figure page 1-1 Absorption coefficient versus frequency . . . . . . . . . . . . . . . . . . . . . . 18 1-2 Underwater channel impulse response (CIR) example . . . . . . . . . . . . . . 19 1-3 Scattering functions obtained at two different conditions . . . . . . . . . . . . . 20 1-4 Normalized CIR evolution over approximately a one-minite period . . . . . . . . 21 2-1 An N (cid:2)M MIMO UAC system . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 2-2 The structure of the LMMSE based soft-input soft-output equalizer . . . . . . . 56 2-3 Simulation averaged over 500 Monte-Carlo trials . . . . . . . . . . . . . . . . . 56 2-4 The structure of the package used in the MACE10 experiment . . . . . . . . . 57 2-5 The structure of the transmitted symbols for the 4(cid:2)12 MIMO BLAST scheme used in MACE10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 2-6 The superimposed modulus of the CIR estimates . . . . . . . . . . . . . . . . . 58 2-7 The effect of resampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 2-8 The relative speed between the transmitter and receiver array given by GPS and estimated during the temporal resampling stage . . . . . . . . . . . . . . . 59 2-9 Doppler frequency evolution of the 1st packet in epoch “E018” obtained by GoSLIM and GoSLIM-V . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 2-10 CIR estimation comparison between GoSLIM and GoSLIM-V . . . . . . . . . . 61 2-11 Grayscale mascot obtained from GoSLIM and GoSLIM-V . . . . . . . . . . . . 62 2-12 Grayscale mascot obtained from RELAX-BLAST and Turbo equalization . . . . 63 2-13 The LLR soft information about the source bits at the output of the decoder . . 63 3-1 The simulation geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 3-2 A generic active sonar scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 3-3 Description of the association problem . . . . . . . . . . . . . . . . . . . . . . . 88 3-4 Range-Doppler images obtained at the first receiver . . . . . . . . . . . . . . . 89 3-5 Range-Doppler images obtained at the second receiver . . . . . . . . . . . . . 90 4-1 Spatial pseudo spectra obtained with a scalar sensor array . . . . . . . . . . . 106 8 4-2 Spatial pseudo spectra obtained with a vector sensor array . . . . . . . . . . . 107 4-3 Spatial pseudo spectra in the case of a weak source . . . . . . . . . . . . . . . 108 4-4 Performance enhancement using RELAX . . . . . . . . . . . . . . . . . . . . . 109 4-5 Empirical failure rate and RMSEs versus SNR . . . . . . . . . . . . . . . . . . 109 9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy UNDERWATER ACOUSTIC SIGNAL PROCESSING AND ITS APPLICATIONS By Kexin Zhao May 2014 Chair: Jian Li Major: Electrical and Computer Engineering Underwater acoustic signal processing has various applications in the underwater sensing systems, including the underwater acoustic communication (UAC) system and the active sonar system. A well-designed UAC system can achieve reliable and high data-rate communication to facilitate the operation of submarines, undersea sensors, and unmanned undersea vehicles (UUVs) while active sonar system demands effective and efficient signal processing techniques to accurately detect, localize, and even track the target of interest. The focus of this dissertation is to use the appropriate signal processing techniques to design such UAC and active sonar systems. For the former, we focus on designing a mobile multi-input multi-output (MIMO) UAC system over double-selective channels subject to both inter-symbol interference and Doppler scaling effects. Temporal resampling is implemented to effectively convert the Doppler scaling effects to Doppler frequency shifts. By simplifying the assumption on the Doppler frequency shifts imposed on the channel taps across all the transmitter and receiver pairs, two sparse channel estimation algorithms, both as an extension of the original sparse learning via iterative minimization (SLIM) method, are proposed for channel estimation. Regarding symbol detection, we employ Turbo equalization and propose a fast implementation of the standard Turbo equalizer for retrieving the transmitted signal. The effectiveness of the considered mobile MIMO UAC scheme is 10