An Information-Theoretic Approach to Waveform Design for Active Incoherent Microwave Imaging

Abstract

Active incoherent imaging is a new microwave imaging method that combines spatial frequency sampling methods with active noise radar techniques. In this paper, we present a method for characterizing image reconstruction fidelity by analyzing the transmitted noise waveforms in terms of the mutual information between them. Different statistical distributions of the transmitted noise are characterized, and we investigate the effect on mutual information and image fidelity in terms of the bandwidth of the received waveforms. We show that Gaussian noise generally outperforms other distributions, and that the mutual information follows the general trend of image fidelity, indicating that mutual information can be a useful metric for designing transmit noise waveforms for active incoherent imaging.