Analysis of Optimal Thresholding Algorithms for Compressed Sensing

Abstract

The optimal $k$-thresholding (OT) and optimal $k$-thresholding pursuit (OTP) are newly introduced frameworks of thresholding techniques for compressed sensing and signal approximation. Such frameworks motivate the practical and efficient algorithms called relaxed optimal $k$-thresholding ($\textrm{ROT}\omega$) and relaxed optimal $k$-thresholding pursuit ($\textrm{ROTP}\omega$) which are developed through the tightest convex relaxations of OT and OTP, where $\omega$ is a prescribed integer number. The preliminary numerical results demonstrated in \cite{Z19} indicate that these approaches can stably reconstruct signals with a wide range of sparsity levels. However, the guaranteed performance of these algorithms with parameter $ \omega \geq 2 $ has not yet established in \cite{Z19}. The purpose of this paper is to show the guaranteed performance of OT and OTP in terms of the restricted isometry property (RIP) of nearly optimal order for the sensing matrix governing the $k$-sparse signal recovery, and to establish the first guaranteed performance result for $\textrm{ROT}\omega$ and $\textrm{ROTP}\omega$ with $ \omega\geq 2.$ In the meantime, we provide a numerical comparison between ROTP$\omega$ and several existing thresholding methods.