Iteratively reweighted least squares for block-sparse recovery

The compressive sensing (CS) theory has shown that sparse signals can be reconstructed exactly from much fewer measurements than traditionally believed. What's more, using ℓ
_p
-norm minimization with p <; 1 can do so with much fewer measurements than with p=1. In this paper, a novel algorithm is proposed for computing local minima of the nonconvex problem in the block-sparse system. A series of experiments are presented to show the remarkable performance of our proposed algorithm in block sparse signal recovery, and compare the recovery ability of this algorithm with the IRLS and BOMP algorithm.