Improved Bounds on the Restricted Isometry Constant for Orthogonal Matching Pursuit
arXiv:1406.4335
Abstract
In this letter, we first construct a counter example to show that for any given positive integer and for any , there always exist a sparse $\x$ and a matrix $\A$ with the restricted isometry constant such that the OMP algorithm fails in iterations. Secondly, we show that even when , the OMP algorithm can also perfectly recover every sparse vector $\x$ from $\y=\A\x$ in iteration. This improves the best existing results which were independently given by Mo et al. and Wang et al.
Electronic Letters, 2013