Adaptive reduced-set matching pursuit for compressed sensing recovery

Abstract

Compressed sensing enables the acquisition of sparse signals at a rate that is much lower than the Nyquist rate. Various greedy recovery algorithms have been proposed to achieve a lower computational complexity compared to the optimal ℓ1 minimization, while maintaining a good reconstruction accuracy. We propose a new greedy recovery algorithm for compressed sensing, called the Adaptive Reduced-set Matching Pursuit (ARMP). Our algorithm achieves higher reconstruction accuracy at a significantly low computational complexity compared to existing greedy recovery algorithms. It is even superior to ℓ1 minimization in terms of the normalized time-error product, a metric that we introduced to measure the trade-off between the reconstruction time and error.