Can I read Support Recovery with Sparsely Sampled Free Random Matrices on EtoBox?

Support Recovery with Sparsely Sampled Free Random Matrices by Tulino, Antonia; Caire, Giuseppe; Verdu', Sergio; Shamai, Shlomo is a scholarly article available to read on EtoBox.

What is Support Recovery with Sparsely Sampled Free Random Matrices about?

Consider a Bernoulli-Gaussian complex $n$-vector whose components are $V_i = X_i B_i$, with $X_i \sim \Cc\Nc(0,\Pc_x)$ and binary $B_i$ mutually independent and iid across $i$. This random $q$-sparse vector is multiplied by a square random matrix $\Um$, and a randomly chosen subset, of average size $n p$, $p \in [0,1]$, of the resulting vector components is then observed in additive Gaussian noise. We extend the scope of conventional noisy compressive sampling models where $\Um$ is typically %A16 the identity or a matrix with iid components, to allow $\Um$ satisfying a certain freeness condition. This class of matrices encompasses Haar matrices and other unitarily invariant matrices. We use the replica method and the decoupling principle of Guo and Verd\'u, as well as a number of information theoretic bounds, to study the input-output mutual information and the support recovery error rate in the limit of $n \to \infty$. We also extend the scope of the large deviation approach of Rangan, Fletcher and Goyal and characterize the performance of a class of estimators encompassing thresholded linear MMSE and $\ell_1$ relaxation.

Author

Tulino, Antonia; Caire, Giuseppe; Verdu', Sergio; Shamai, Shlomo

Published

2012

Language

EN