Robust signal recovery using Bayesian compressed sensing based on Lomax prior

Abstract

Recently published research shows that Lomax distribution exhibits compressibility in Lorentz curves. In this paper, we address the problem of signal reconstruction in the high noise level and phase error environments in a Bayesian framework of Lomax prior distribution. Furthermore, from the perspective of improving sparsity and compressibility of the signal constraints, a novel reconstruction model deducted from Lomax-prior-based Bayesian compressed sensing (LomaxCS) is proposed. The LomaxCS improves the accuracy of existing Bayesian compressed sensing signal reconstruction methods and enhances the robustness against Gauss noise and phase errors. Compared with the conventional models, the proposed LomaxCS model still reveals the general profile of the signal in the worst conditions. The experimental results demonstrate that the proposed algorithm can achieve substantial improvements in terms of recovering signal quality and robustness; meanwhile, it brings an evident application prospect.