Title :
Recovery Guarantees for Rank Aware Pursuits
Author :
Blanchard, Jeffrey D. ; Davies, Mike E.
Author_Institution :
Dept. of Math. & Stat, Grinnell Coll., Grinnell, IA, USA
fDate :
7/1/2012 12:00:00 AM
Abstract :
This letter considers sufficient conditions for sparse recovery in the sparse multiple measurement vector (MMV) problem for some recently proposed rank aware greedy algorithms. Specifically we consider the compressed sensing framework with Gaussian random measurement matrices and show that the rank of the measurement matrix in the noiseless sparse MMV problem allows such algorithms to reduce the effect of the logn term that is present in traditional OMP recovery.
Keywords :
computational complexity; greedy algorithms; matrix algebra; signal representation; Gaussian random measurement matrices; compressed sensing framework; noiseless sparse MMV problem; rank aware greedy algorithms; rank aware pursuits; signal representations; sparse multiple measurement vector problem; sparse recovery; traditional OMP recovery; Algorithm design and analysis; Joints; Matching pursuit algorithms; Multiple signal classification; Sparse matrices; Vectors; Greedy algorithm; multiple measurement vectors; orthogonal matching pursuit; rank;
Journal_Title :
Signal Processing Letters, IEEE
DOI :
10.1109/LSP.2012.2199752
