40
167

Information-theoretic limits on sparse signal recovery: Dense versus sparse measurement matrices

Abstract

We study the information-theoretic limits of exactly recovering the support of a sparse signal using noisy projections defined by various classes of measurement matrices. Our analysis is high-dimensional in nature, in which the number of observations nn, the ambient signal dimension pp, and the signal sparsity kk are all allowed to tend to infinity in a general manner. This paper makes two novel contributions. First, we provide sharper necessary conditions for exact support recovery using general (non-Gaussian) dense measurement matrices. Combined with previously known sufficient conditions, this result yields sharp characterizations of when the optimal decoder can recover a signal for various scalings of the sparsity kk and sample size nn, including the important special case of linear sparsity (k=Θ(p)k = \Theta(p)) using a linear scaling of observations (n=Θ(p)n = \Theta(p)). Our second contribution is to prove necessary conditions on the number of observations nn required for asymptotically reliable recovery using a class of γ\gamma-sparsified measurement matrices, where the measurement sparsity γ(n,p,k)(0,1]\gamma(n, p, k) \in (0,1] corresponds to the fraction of non-zero entries per row. Our analysis allows general scaling of the quadruplet (n,p,k,γ)(n, p, k, \gamma), and reveals three different regimes, corresponding to whether measurement sparsity has no effect, a minor effect, or a dramatic effect on the information-theoretic limits of the subset recovery problem.

View on arXiv
Comments on this paper