Greedy pursuit algorithms for sparse signal reconstruction in the case of impulsive noise

Abstract

This work addresses sparse signal reconstruction problem in the presence of impulsive noise based on greedy algorithms. The impulsive noise, in this paper, is modeled by an unknown sparse vector so that the impulsive suppression problem is transformed to an estimation problem. To accurately recover the signal of interest (SOI) and suppress the noise, a joint estimation is devised to simultaneously perform the noise suppression and the SOI recovery. To efficiently solve the problem, greedy algorithm based iterative procedure is developed in which both the SOI and the impulsive noise are estimated utilizing their sparse structures. Simulation results demonstrate the superior performance of the joint greedy estimation algorithm.