Analysis and design of algorithms for compressive sensing based noise radar systems

We study the compressive radar imaging problem from the perspective of statistical estimation. The goal of this paper is to characterize the estimation error. Conventional radar estimation and detection techniques are characterized by concrete performance guarantees which relate directly to practical systems. The state evolution approach applied to compressive sensing is particularly useful for such analysis. We emphasize the importance of the uniform norm of the estimation error for radar imaging. In the second part of the paper, we propose a weighted compressive sampling scheme for noise radar imaging that utilizes prior information about the target scene. The weights are obtained using the mutual information estimation between target echoes and the transmitted signals with an energy constraint.