Compressed sensing signal recovery via A* Orthogonal Matching Pursuit

2011 ◽  
Author(s):  
Nazim Burak Karahanoglu ◽  
Hakan Erdogan
Keyword(s):  
Compressed Sensing ◽  
Matching Pursuit ◽  
Orthogonal Matching Pursuit ◽  
Signal Recovery

An Optimised Signal Reconstruction Algorithm for Compressed Sensing

2013 ◽  
Vol 333-335 ◽  
pp. 567-571
Author(s):  
Zhao Shan Wang ◽  
Shan Xiang Lv ◽  
Jiu Chao Feng ◽  
Yan Sheng ◽  
Zhong Liang Wu ◽  
...  
Keyword(s):  
Compressed Sensing ◽  
Small Proportion ◽  
Matching Pursuit ◽  
Signal Reconstruction ◽  
Reconstruction Algorithm ◽  
Orthogonal Matching Pursuit ◽  
Signal Recovery ◽  
Successful Recovery ◽  
Stagewise Orthogonal Matching Pursuit ◽  
Convergence Of Algorithm

Signal recovery is a key issue in compressed sensing field. A new greedy reconstruction algorithm termed Optimised Stagewise Orthogonal Matching Pursuit (OSOMP) is proposed, which is an improved version for Stagewise Orthogonal Matching Pursuit (StOMP). In preselection step, OSOMP chooses several coordinates with a calculated threshold to accelerate the convergence of algorithm. In following pruning step, a small proportion of selected coordinates are discarded according to the amplitude of estimated signal, thus most false discovered coordinates can be swept away. Experimental results show that in OSOMP, the scale of estimated support can be controlled very well, and the successful recovery rate is also much higher than that in StOMP.

Least Support Orthogonal Matching Pursuit Algorithm With Prior Information

2014 ◽  
Vol 6 (2) ◽  
pp. 111-134 ◽  
Author(s):  
Israa Sh. Tawfic ◽  
Sema Koc Kayhan
Keyword(s):  
Numerical Experiments ◽  
Prior Information ◽  
Matching Pursuit ◽  
Recovery Process ◽  
Orthogonal Matching Pursuit ◽  
Sparse Signal Recovery ◽  
Sparse Signals ◽  
Signal Recovery ◽  
Reconstruction Performance ◽  
Matching Pursuit Algorithm

Abstract This paper proposes a new fast matching pursuit technique named Partially Known Least Support Orthogonal Matching Pursuit (PKLS-OMP) which utilizes partially known support as a prior knowledge to reconstruct sparse signals from a limited number of its linear projections. The PKLS-OMP algorithm chooses optimum least part of the support at each iteration without need to test each candidate independently and incorporates prior signal information in the recovery process. We also derive sufficient condition for stable sparse signal recovery with the partially known support. Result shows that inclusion of prior information weakens the condition on the sensing matrices and needs fewer samples for successful reconstruction. Numerical experiments demonstrate that PKLS-OMP performs well compared to existing algorithms both in terms of reconstruction performance and execution time.

Sign in / Sign up

Export Citation Format

Share Document