A Block-Sparse Signal Recovery Method Based on Adaptive Filtering Framework

In this paper, we present a new method for block-sparse signals reconstruction under the framework of adaptive filtering. Firstly, the block-sparse structure is exploited and introduced into the adaptive filtering framework. Furthermore, the mixed l2-lo norm is employed to represent this structure. Then, the idea of smoothed l0 (SLO) is utilized to approximate the mixed norm. Thirdly, a modified zero-attractor term is developed to further improve the recover performance. Finally, some experiments are carried out to demonstrate the effectiveness of the proposed approach.