Robotic Navigation and Mapping with Radar -- Contents -- Preface -- Acknowledgments -- Acronyms -- Nomenclature -- Chapter 1 Introduction -- 1.1 Isn't Navigation and Mapping with Radar Solved? -- 1.1.1 Applying Missile/Aircraft Guidance Technologies to Robotic Vehicles -- 1.1.2 Placing Autonomous Navigation of Robotic Vehicles into Perspective -- 1.2 Why Radar in Robotics? Motivation -- 1.3 The Direction of Radar-based Robotics Research -- 1.3.1 Mining Applications -- 1.3.2 Intelligent Transportation System Applications -- 1.3.3 Land-Based SLAM Applications -- 1.3.4 Coastal Marine Applications -- 1.4 Structure of the Book -- References -- PART I: Fundamentals of Radar and Robotic Navigation -- Chapter 2 A Brief Overview of Radar Fundamentals -- 2.1 Introduction -- 2.2 Radar Measurements -- 2.3 The Radar Equation -- 2.4 Radar Signal Attenuation -- 2.5 Measurement Power Compression and Range Compensation -- 2.5.1 Logarithmic Compression -- 2.5.2 Range Compensation -- 2.5.3 Logarithmic Compression and Range Compensation During Target Absence -- 2.5.4 Logarithmic Compression and Range Compensation During Target Presence -- 2.6 Radar-Range Measurement Techniques -- 2.6.1 Time-of-Flight (TOF) Pulsed Radar -- 2.6.2 Frequency Modulated Continuous Wave (FMCW) Radar -- 2.6.2.2 Doppler Measurements -- 2.6.2.3 Multiple Line-of-Sight Targets -- 2.7 Sources of Uncertainty in Radar -- 2.7.1 Sources of Uncertainty Common to All Radar Types -- 2.8 Uncertainty Specific to TOF and FMCW Radar -- 2.8.1 Uncertainty in TOF Radars -- 2.8.2 Uncertainty in FMCW Radars -- 2.9 Polar to Cartesian Data Transformation -- 2.9.1 Nearest Neighbor Polar to Cartesian Data Conversion -- 2.9.2 Weighted Polar to Cartesian Data Conversion -- 2.10 Summary -- 2.11 Bibliographical Remarks -- 2.11.1 Extensions to the Radar Equation -- 2.11.2 Signal Propagation/Attenuation.

2.11.3 Range Measurement Methods -- 2.11.4 Uncertainty in Radar -- References -- Chapter 3 An Introduction to Detection Theory -- 3.1 Introduction -- 3.2 Concepts of Detection Theory -- 3.3 Different Approaches to Target Detection -- 3.3.1 Non-adaptive Detection -- 3.3.2 Hypothesis Free Modeling -- 3.3.3 Stochastic-Based Adaptive Detection -- 3.4 Detection Theory with Known Noise Statistics -- 3.4.1 Constant CFARPfa with Known Noise Statistics -- 3.4.2 Probability of Detection CFARPD with Known Noise Statistics -- 3.4.3 Probabilities of Missed Detection CFARPMD and Noise CFARPn with Known Noise Statistics -- 3.5 Detection with Unknown Noise Statistics-Adaptive CFAR Processors -- 3.5.1 Cell Averaging-CA-CFAR Processors -- 3.5.2 Ordered Statistics-OS-CFAR Processors -- 3.6 Summary -- 3.7 Bibliographical Remarks -- References -- Chapter 4 Robotic Navigation and Mapping -- 4.1 Introduction -- 4.2 General Bayesian SLAM-The Joint Problem -- 4.2.1 Vehicle State Representation -- 4.2.2 Map Representation -- 4.3 Solving Robot Mapping and Localization Individually -- 4.3.1 Probabilistic Robotic Mapping -- 4.3.2 Probabilistic Robotic Localization -- 4.4 Popular Robotic Mapping Solutions -- 4.4.1 Grid-Based Robotic Mapping (GBRM) -- 4.4.2 Feature-Based Robotic Mapping (FBRM) -- 4.5 Relating Sensor Measurements to Robotic Mapping and SLAM -- 4.5.1 Relating the Spatial Measurement Interpretation to the Mapping/SLAM State -- 4.5.2 Relating the Detection Measurement Interpretation to the Mapping/SLAM State -- 4.6 Popular FB-SLAM Solutions -- 4.6.1 Bayesian FB-SLAM-Approximate Gaussian Solutions -- 4.6.2 Feature Association -- 4.6.3 Bayesian FB-SLAM-Approximate Particle Solutions -- 4.6.4 A Factorized Solution to SLAM (FastSLAM) -- 4.6.5 Multi-Hypothesis (MH) FastSLAM -- 4.6.6 General Comments on Vector-Based FB SLAM -- 4.7 FBRM and SLAM with Random Finite Sets.

4.7.1 Motivation: Why Random Finite Sets -- 4.7.2 RFS Representations of State and Detected Features -- 4.7.3 Bayesian Formulation with a Finite Set Feature Map -- 4.7.4 The Probability Hypothesis Density (PHD) Estimator -- 4.7.5 The PHD Filter -- 4.8 SLAM and FBRM Performance Metrics -- 4.8.1 Vehicle State Estimate Evaluation -- 4.8.2 Map Estimate Evaluation -- 4.8.3 Evaluation of FBRM and SLAM with the Second Order Wasserstein Metric -- 4.9 Summary -- 4.10 Bibliographical Remarks -- 4.10.1 Grid-Based Robotic Mapping (GBRM -- 4.10.2 Gaussian Approximations to Bayes Theorem -- 4.10.3 Non-Parametric Approximations to Bayesian FB-SLAM -- 4.10.4 Other Approximations to Bayesian FB-SLAM -- 4.10.5 Feature Association and Management -- 4.10.6 Random Finite Sets (RFSs) -- 4.10.7 SLAM and FBRM Evaluation Metrics -- References -- Part II: Radar Modeling and Scan Integration -- Chapter 5 Predicting and Simulating FMCW Radar Measurements -- 5.1 Introduction -- 5.2 FMCW Radar Detection in the Presence of Noise -- 5.3 Noise Distributions During Target Absence and Presence -- 5.3.1 Received Power Noise Estimation -- 5.3.2 Range Noise Estimation -- 5.4 Predicting Radar Measurements -- 5.4.1 A-Scope Prediction Based on Expected Target RCS and Range -- 5.4.2 A-Scope Prediction Based on Robot Motion -- 5.5 Quantitative Comparison of Predicted and Actual Measurements -- 5.6 A-scope Prediction Results -- 5.6.1 Single Bearing A-Scope Prediction -- 5.6.2 360° Scan Multiple A-Scope Prediction, Based on Robot Motion -- 5.7 Summary -- 5.8 Bibliographical Remarks -- References -- Chapter 6 Reducing Detection Errors and Noise with Multiple Radar Scans -- 6.1 Introduction -- 6.2 Radar Data in an Urban Environment -- 6.2.1 Landmark Detection with Single Scan CA-CFAR -- 6.3 Classical Scan Integration Methods -- 6.3.1 Coherent and Noncoherent Integration.

6.3.2 Binary Integration Detection -- 6.4 Integration Based on Target Presence Probability (TPP) Estimation -- 6.5.4 Numerical Method for Determining T TPP (αp ,l ) and TPP -- 6.5.1 TPP Response to Noise: TPP -- 6.5.2 TPP Response to a Landmark and Noise: TPP -- 6.5.3 Choice of αp, TTPP (αp, l ) and l -- 6.6 A Comparison of Scan Integration Methods -- 6.7 A Note on Multi-Path Reflections -- 6.8 TPP Integration of Radar in an Urban Environment -- 6.8.1 Qualitative Assessment of TPP Applied to A-Scope Information -- 6.8.2 Quantitative Assessment of TPP Applied to Complete Scans -- 6.8.3 A Qualitative Assessment of an Entire Parking Lot Scene -- 6.9 Recursive A-Scope Noise Reduction -- 6.9.1 Single A-Scope Noise Subtraction -- 6.9.2 Multiple A-Scope-Complete Scan Noise Subtraction -- 6.10 Summary -- 6.11 Bibliographical Remarks -- References -- Part III: IRobotic Mapping with KnownVehicle Location -- Chapter 7 Grid-Based Robotic Mapping with Detection Likelihood Filtering -- 7.1 Introduction -- 7.2 The Grid-Based Robotic Mapping (GBRM) Problem -- 7.2.1 GBRM Based on Range Measurements -- 7.2.2 GBRM with Detection Measurements -- 7.2.3 Detection versus Range Measurement Models -- 7.3 Mapping with Unknown Measurement Likelihoods -- 7.3.1 Data Format -- 7.3.2 GBRM Algorithm Overview -- 7.3.3 Constant False Alarm Rate (CFAR) Detector -- 7.3.4 Map Occupancy and Detection Likelihood Estimator -- 7.3.5 Incorporation of the OS-CFAR Processor -- 7.4 GBRM-ML Particle Filter Implementation -- 7.5 Comparisons of Detection and Spatial-Based GBRM -- 7.5.1 Dataset 1: Synthetic Data, Single Landmark -- 7.5.2 Dataset 2: Real Experiments in the Parking Lot Environment -- 7.5.3 Dataset 3: A Campus Environment -- 7.6 Summary -- 7.7 Bibliographical Remarks -- References -- Chapter 8 Feature-Based Robotic Mapping with Random Finite Sets -- 8.1 Introduction.

8.2 The Probability Hypothesis Density (PHD)-FBRM Filter -- 8.3 PHD-FBRM Filter Implementation -- 8.3.1 The FBRM New Feature Proposal Strategy -- 8.3.2 Gaussian Management and State Estimation -- 8.3.3 GMM-PHD-FBRM Pseudo Code -- 8.4 PHD-FBRM Computational Complexity -- 8.5 Analysis of the PHD-FBRM Filter -- 8.6 Summary -- 8.7 Bibliographical Remarks -- References -- Part IV Simultaneous Localization and Mapping -- Chapter 9 Radar-Based SLAM with Random Finite Sets -- 9.1 Introduction -- 9.2 Slam with the PHD Filter -- 9.2.1 The Factorized RFS-SLAM Recursion -- 9.2.2 PHD Mapping-Rao-Blackwellization -- 9.2.3 PHD-SLAM -- 9.3 Implementing the RB-PHD-SLAM Filter -- 9.3.1 PHD Mapping-Implementation -- 9.3.2 The Vehicle Trajectory-Implementation -- 9.3.3 Estimating the Map -- 9.3.4 GMM-PHD-SLAM Pseudo Code -- 9.4 RB-PHD-Slam Computational Complexity -- 9.5 Radar-Based Comparisons of RFS and Vector-Based SLAM -- 9.6 Summary -- 9.7 Bibliographical Remarks -- References -- Chapter 10 X-Band Radar-Based SLAM in an Off-Shore Environment -- 10.1 Introduction -- 10.2 The ASC and the Coastal Environment -- 10.3 Marine Radar Feature Extraction -- 10.3.1 Adaptive Coastal Feature Detection-OS-CFAR -- 10.3.2 Image-Based Smoothing-Gaussian Filtering -- 10.3.3 Image-Based Thresholding -- 10.3.4 Image-Based Clustering -- 10.3.5 Feature Labeling -- 10.4 The Marine-Based SLAM Algorithms -- 10.4.1 The ASC Process Model -- 10.4.2 RFS SLAM with the PHD Filter -- 10.4.3 NN-EKF-SLAM Implementation -- 10.4.4 Multi-Hypothesis (MH) FastSLAM Implementation -- 10.5 Comparisons of SLAM Concepts at Sea -- 10.5.1 SLAM Trial 1-Comparing PHD and NN-EKF-SLAM -- 10.5.2 SLAM Trial 2-Comparing RB-PHD-SLAM and MH-FastSLAM -- 10.6 Summary -- 10.7 Bibliographical Remarks -- References -- Appendix A: The Navtech FMCW MMW Radar Specifications.

Appendix B: Derivation of g(Zk