Information-theoretic waveform design for MIMO radar detection in range-spread clutter

摘要

This paper addresses the problem of waveform design for multiple-input-multiple-output (MIMO) radar detection in the presence of range-spread clutter. Owing to the intractability of waveform design based on maximizing the detection probability, the design metric employs the relative entropy between the probability density functions of the observations under the two hypotheses (viz., the target is present/absent). We consider several practical constraints on the waveforms, including energy constraint, peak-to-average-power ratio constraint, similarity constraint, and both constant-modulus and similarity constraints. We propose a minorization-maximization approach to tackle the non-convex waveform design problem. We show that the proposed algorithm has guaranteed and fast convergence of objective values. Finally, numerical examples are provided to demonstrate the effectiveness of the proposed algorithm.