Compressed Sensing for High-Speed Maneuvering Radar Targets: A Fundamental Analysis

This paper investigates the fundamental theoretical limits of applying compressed sensing (CS) to high-speed maneuvering radar target detection and parameter estimation.Through rigorous mathematical analysis, we examine the signalcharacteristics, dictionary design constraints, and reconstruction performance bounds for targets exhibiting complex motion including acceleration and jerk. Our analysis reveals critical limitations in achieving the restricted isometry property (RIP) for realistic parameter spaces, with dictionary coherence exceeding practical thresholds. We derive Cramér-Rao lower bounds for parameter estimation and demonstrate that while CS offers theoretical computational advantages through separable dictionary structures, the inherent signal characteristics of maneuvering targets violate key CS assumptions. The finding suggest that compressed sensing may not be the optimal framework for this class of radar problems, providing theoretical justification for alternative approaches.