MIMO radar using sparse sensing: A weighted sparse Bayesian learning (WSBL) approach

Abstract

A colocated Multiple-Input Multiple-Output (MIMO) radar system is studied, in which the receive antennas implement sparse sensing and then forward their compressively obtained samples to a fusion center. Assuming sparsely distributed targets in the direction-of-arrival (DOA) space, the fusion center can estimate the targets by formulating and solving a sparse signal recovery problem. In this paper, we propose a weighted Sparse Bayesian Learning (WSBL) approach for target DOA estimation. Using a low resolution estimate of the sparse vector, the proposed approach assigns different weights to different entries of the sparse vector, giving more importance to some entries over others. Subsequently, the weighted sparse signal recovery problem is solved along the lines of the Sparse Bayesian Learning (SBL) framework. The proposed approach shows robustness for increased number of sources, and lower SNR as compared to SBL and the Dantzig selector approach.