Low-Complexity Direction Finding Method for MIMO Radar Based on Compressive Sensing

Abstract

We propose a direction finding method for multiple-input multiple-output (MIMO) radar by using sparse sensing in low computational cost. Since the targets are sparsely distributed in the space, we employ the compressive sensing technique to reduce the sampling rate. Based on the compressive sensing, we utilize the convex combination to approximate the real target parameters in MIMO radar. We formulate an optimization problem for sparse vector recovery and off-grid mismatch estimation, which involves four set of variables. We employ the alternating direction method of multipliers approach to fast solve this optimization problem. In each iteration of sparse recovery, four subproblems are alternately optimized over only one of four set of parameters where each subproblem has a closed-form solution. With the recovered sparse vector and the estimated off-grid mismatch, we develop a grid adjustment method to accurately resolve the directions of targets by iteratively deleting the redundant grid points. Numerical simulations indicate that the proposed method is able to achieve accurate signal recovery, improved estimation accuracy and reduced computational cost.