MMV Batch Look Ahead Orthogonal Matching Pursuit (M-BLAOMP) Algorithm for Joint Sparse Recovery.

For solving the problems of multiple measurement vectors, MMV Batch Look Ahead Orthogonal Matching Pursuit (M-BLAOMP) algorithm is proposed based on the M-OMP algorithm to improve the recovery performance. We use a look ahead strategy to examine the effect of the selected atom on the final residual and output a set of support indexes as the candidate support set accordingly. Operation that takes the intersection of the potential support set and the candidate support set can make the algorithm converge quickly. Experimental results show that the M-BLAOMP algorithm has a significant improvement in recovery performance.