Modified Papoulis-Gerchberg algorithm for sparse signal recovery

TL;DR

This paper introduces a modified iterative thresholding algorithm inspired by Papoulis-Gerchberg for recovering sparse signals from limited observations, demonstrating competitive performance through experimental evaluation.

Contribution

It proposes a novel variation of the Papoulis-Gerchberg algorithm tailored for sparse signal recovery, with a focus on iterative thresholding techniques.

Findings

The algorithm effectively recovers sparse signals from few observations.

Experimental results show competitive performance compared to existing methods.

The iterative process resembles thresholded Landweber iterations.

Abstract

Motivated by the well-known Papoulis-Gerchberg algorithm, an iterative thresholding algorithm for recovery of sparse signals from few observations is proposed. The sequence of iterates turns out to be similar to that of the thresholded Landweber iterations, although not the same. The performance of the proposed algorithm is experimentally evaluated and compared to other state-of-the-art methods.