A Novel Deterministic Sensing Matrix Based on Kasami Codes for Cluster Structured Sparse Signals

Abstract

Cluster structured compressive sensing is a new direction of compressive sensing, dealing with cluster structured sparse signals. In this paper, we propose a sensing matrix based on Kasami codes for CSS signals. The Kasami codes have been the subject of several constructions. Our idea is to make these constructions suitable to CSS signals. The proposed matrix, gives more intention to the clusters. Simulation results show the superior performance of our matrix. In that, it gives the highest rate of exact recovery. Moreover, the deterministic aspect of our matrix makes it more suitable to be implemented on hardware.