A Robust Pre-processing Approach of Array Samples with Hankel Matrix Completion

Abstract

In this paper, we study the problem of recovering array samples when a few array samples are unavailable, and provide a robust pre-processing approach of recovering unavailable array samples with Hankel matrix completion. First, Alternating Direction Methods (ADM) for solving nuclear norm minimization problem with Hankel structure was discussed. Second, stability of Hankel matrix reconstruction is investigated in view of l*-Constrained Minimal Singular Value (l*-CMSV). Finally, simulation experiments illustrate that the proposed method's are robust for complex Gaussian noise.