Information-Theoretic Methods for Waveform Design in Multistatic Radar Imaging

In this chapter, we present a waveform optimization procedure for multistatic radar imaging that is based on the mutual information (MI) maximization criterion. A multistatic radar scenario is considered for sparse and random sensor positions. It is seen that natural scenes exhibit a natural structure over which a prior distribution can be placed. Specifically, a compound Gaussian (CG) distribution prior can be placed on the wavelet coefficients on the scene. We present a novel waveform design technique with respect to the MI criterion based on CG scene statistics when it is projected onto a sparse basis. Upper and lower bounds are presented that can be optimized using discrete techniques to find an optimal waveform within the context of multistatic radar.