Sparse Recovery for Compressive Sensing via Weighted Lp-q Model

Abstract

In this paper, we study weighted Lp????q minimization model which comprises non-smooth, non-convex and non-Lipschitz quasi-norm Lp(0 < p 1) and Lq(1 < q 2) for recovering sparse signals. Based on the restricted isometry property(RIP) condition, we obtain exact sparse signal recovery result. We also obtain the theoretical bound for the weighted Lp????q minimization model when measurements are depraved by the noises.