Mismatched sparse denoiser requires overestimating the support length

A well-known result [1, Lemma 3.4] states that, without noise, it is better to overestimate the support of a sparse signal, since, if the estimated support includes the true support, the reconstruction is perfect. In this paper, we investigate whether this result holds also in the presence of noise. First, we derive the covariance matrix of the signal estimate when the observation matrix is Gaussian, generalizing existing results. Then, we show that, even in the noisy case, overestimating the support length is the preferred solution, as the error incurred by missing some signal components dominates the overall error variance. Finally, an upper bound of the estimated support length is provided to avoid excessive noise amplification.

[1]  David L Donoho,et al.  Compressed sensing , 2006, IEEE Transactions on Information Theory.

[2]  J. Tropp,et al.  CoSaMP: Iterative signal recovery from incomplete and inaccurate samples , 2008, Commun. ACM.

[3]  Robb J. Muirhead Wiley Series in Probability and Statistics , 1982 .

[4]  Galen Reeves,et al.  The Sampling Rate-Distortion Tradeoff for Sparsity Pattern Recovery in Compressed Sensing , 2010, IEEE Transactions on Information Theory.

[5]  Olgica Milenkovic,et al.  Subspace Pursuit for Compressive Sensing Signal Reconstruction , 2008, IEEE Transactions on Information Theory.

[6]  Galen Reeves,et al.  Approximate Sparsity Pattern Recovery: Information-Theoretic Lower Bounds , 2010, IEEE Transactions on Information Theory.

[7]  Joel A. Tropp,et al.  Signal Recovery From Random Measurements Via Orthogonal Matching Pursuit , 2007, IEEE Transactions on Information Theory.

[8]  Enrico Magli,et al.  Exact performance analysis of the oracle receiver for compressed sensing reconstruction , 2014, 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[9]  Sundeep Rangan,et al.  Orthogonal Matching Pursuit: A Brownian Motion Analysis , 2011, IEEE Transactions on Signal Processing.

[10]  Volkan Cevher,et al.  Compressible Distributions for High-Dimensional Statistics , 2011, IEEE Transactions on Information Theory.

[11]  Michael Elad,et al.  Sparse and Redundant Representations - From Theory to Applications in Signal and Image Processing , 2010 .

[12]  R. DeVore,et al.  A Simple Proof of the Restricted Isometry Property for Random Matrices , 2008 .

[13]  Deanna Needell,et al.  CoSaMP: Iterative signal recovery from incomplete and inaccurate samples , 2008, ArXiv.

[14]  Emmanuel J. Candès,et al.  Near-Optimal Signal Recovery From Random Projections: Universal Encoding Strategies? , 2004, IEEE Transactions on Information Theory.

[15]  Enrico Magli,et al.  Operational Rate-Distortion Performance of Single-Source and Distributed Compressed Sensing , 2014, IEEE Transactions on Communications.

[16]  D. Rosen Moments for the inverted Wishart distribution , 1988 .

[17]  R. Muirhead Aspects of Multivariate Statistical Theory , 1982, Wiley Series in Probability and Statistics.