Greedy Pursuits Assisted Basis Pursuit for reconstruction of joint-sparse signals

摘要: Abstract Distributed Compressive Sensing (DCS) is an extension of compressive sensing from single measurement vector problem to Multiple Measurement Vectors (MMV) problem. In DCS, several reconstruction algorithms have been proposed reconstruct the joint-sparse signal ensemble. However, most them are designed for ensemble sharing common support. Since assumption sparsity pattern very restrictive, we more interested in containing both and innovation components. With a goal proposing MMV-type algorithm that robust outliers (absence pattern), propose Greedy Pursuits Assisted Basis Pursuit (GPABP-MMV). It employs modified basis pursuit MMV versions multiple greedy pursuits. We also formulate exact conditions error bound GPABP-MMV. GPABP-MMV suitable variety applications including time-sequence video frames, ECG signals, etc.