ML-Based Block Sparse Recovery for distributed MIMO Radars in Clutter Environments

In distributed MIMO radars, block sparse recovery methods can be used to estimate the multiple targets parameters while the sampling rate at each receiver is reduced by exploiting Compressive Sensing (CS). However, the performances of these methods severely degrade in strong clutter environments. In this paper, we propose a Maximum Likelihood (ML) based block sparse recovery scheme called ML-BSR for target localization in these environments. The ML optimization requires a very high computational complexity particularly for the localization of an unknown number of targets. However, the complexity of the proposed scheme is acceptable. The proposed scheme has an accurate performance in clutter environments even with reduced the sampling rate at the receivers.