ADAPTIVE RADAR DETECTION : model-based, data-driven and hybrid approaches

This book shows you how to adopt data-driven techniques for the problem of radar detection, both per se and in combination with model-based approaches. In particular, the focus is on space-time adaptive target detection against a background of interference consisting of clutter, possible jammers, and noise. It is a handy, concise reference for many classic (model-based) adaptive radar detection schemes as well as the most popular machine learning techniques (including deep neural networks) and helps you identify suitable data-driven approaches for radar detection and the main related issues. You’ll learn how data-driven tools relate to, and can be coupled or hybridized with, traditional adaptive detection statistics; understand fundamental concepts, schemes, and algorithms from statistical learning, classification, and neural networks domains. The book also walks you through how these concepts and schemes have been adapted for the problem of radar detection in the literature and provides you with a methodological guide for the design, illustrating different possible strategies. You’ll be equipped to develop a unified view, under which you can exploit the new possibilities of the data-driven approach even using simulated data. This book is an excellent resource for Radar professionals and industrial researchers, postgraduate students in electrical engineering and the academic community. Artech House Radar Series Adaptive Radar Detection Model-Based, Data-Driven, and Hybrid Approaches 2 Contents 6 Preface 10 Acknowledgments 16 1 Model-Based Adaptive Radar Detection 18 1.1 Introduction to Radar Processing 18 1.1.1 Generalities and Basic Terminology of Coherent Radars 19 1.1.2 Array Processing and Space-Time Adaptive Processing 22 1.1.3 Target Detection and Performance Metrics 25 1.2 Unstructured Signal in White Noise 26 1.2.1 Old but Gold: Basic Signal Detection and the Energy Detector 26 1.2.2 The Neyman–Pearson Approach 28 1.2.3 Adaptive CFAR Detection 30 1.2.4 Correlated Signal Model in White Noise 32 1.3 Structured Signal in White Noise 35 1.3.1 Detection of a Structured Signal in White Noise and Matched Filter 35 1.3.2 Generalized Likelihood Ratio Test 37 1.3.3 Detection of an Unknown Rank-One Signal in White Noise 41 1.3.4 Steering Vector Known up to a Parameter and Doppler Processing 42 1.4 Adaptive Detection in Colored Noise 42 1.4.1 One-Step, Two-Step, and Decoupled Processing 44 1.4.2 General Hypothesis Testing Problem via GLRT: A Comparison 45 1.4.3 Behavior under Mismatched Conditions: Robustness vs Selectivity 48 1.4.4 Model-Based Design of Adaptive Detectors 50 1.5 Summary 59 References 60 2 Classification Problems and Data-Driven Tools 66 2.1 General Decision Problems and Classification 66 2.1.1 M-ary Decision Problems 67 2.1.2 Classifiers and Decision Regions 72 2.1.3 Binary Classification vs Radar Detection 78 2.1.4 Signal Representation and Universal Approximation 81 2.2 Learning Approaches and Classification Algorithms 83 2.2.1 Statistical Learning 83 2.2.2 Bias-Variance Trade-Off 88 2.3 Data-Driven Classifiers 89 2.3.1 k-Nearest Neighbors 90 2.3.2 Linear Methods for Dimensionality Reduction and Classification 92 2.3.3 Support Vector Machine and Kernel Methods 94 2.3.4 Decision Trees and Random Forests 98 2.3.5 Other Machine Learning Tools 101 2.4 Neural Networks and Deep Learning 102 2.4.1 Multilayer Perceptron 103 2.4.2 Feature Engineering vs Feature Learning 105 2.4.3 Deep Learning 106 2.5 Summary 110 References 110 3 Radar Applications of Machine Learning 114 3.1 Data-Driven Radar Applications 114 3.2 Classification of Communication and Radar Signals 117 3.2.1 Automatic Modulation Recognition and Physical-Layer Applications 117 3.2.2 Datasets and Experimentation 119 3.2.3 Classification of Radar Signals and Radiation Sources 124 3.3 Detection Based on Supervised Machine Learning 126 3.3.1 SVM-Based Detection with Controlled PFA 127 3.3.2 Decision Tree-Based Detection with Controlled PFA 128 3.3.3 Revisiting the Neyman–Pearson Approach 129 3.3.4 SVM and NN for CFAR Processing 131 3.3.5 Feature Spaces with (Generalized) CFAR Property 134 3.3.6 Deep Learning Based Detection 137 3.4 Other Approaches 140 3.4.1 Unsupervised Learning and Anomaly Detection 140 3.4.2 Reinforcement Learning 142 3.5 Summary 143 References 144 4 Hybrid Model-Based and Data-Driven Detection 154 4.1 Concept Drift, Retraining, and Adaptiveness 154 4.2 Hybridization Approaches 156 4.2.1 Different Dimensions of Hybridization 156 4.2.2 Hybrid Model-Based and Data-Driven Ideas in Signal Processing and Communications 157 4.3 Feature Spaces Based onWell-Known Statistics or Raw Data 159 4.3.1 Nonparametric Learning: k-Nearest Neighbor 159 4.3.2 Quasi-Whitened Raw Data as Feature Vector 161 4.3.3 Well-Known CFAR Statistics as a Feature Vector 164 4.4 Rethinking Model-Based Detection in a CFAR Feature Space 168 4.4.1 Maximal Invariant Feature Space 168 4.4.2 Characterizing Model-Based Detectors in CFAR-FP 170 4.4.3 Design Strategies in the CFAR-FP 175 4.5 Summary 176 References 177 5 Theories, Interpretability, and Other Open Issues 182 5.1 Challenges in Machine Learning 182 5.2 Theories for (Deep) Neural Networks 184 5.2.1 Network Structures and Unrolling 185 5.2.2 Information Theory, Coding, and Sparse Representation 188 5.2.3 Universal Mapping, Expressiveness, and Generalization 189 5.2.4 Overparametrized Interpolation, Reproducing Kernel Hilbert Spaces, and Double Descent 193 5.2.5 Mathematics of Deep Learning, Statistical Mechanics, and Signal Processing 197 5.3 Open Issues 198 5.3.1 Adversarial Attacks 198 5.3.2 Stability, Efficiency, and Interpretability 199 5.3.3 Visualization 201 5.3.4 Sustainability, Marginal Return, and Patentability 202 5.4 Summary 204 References 205 List of Acronyms 212 List of Symbols 216 About the Author 220 Index 222 Radar;,Adaptive,radar;,Detection;,Artech,House;,978-1-63081-900-2 Radar,Adaptive radar,Detection,Artech House,978-1-63081-900-2