Name
Abstract | ||
|---|---|---|
This paper addresses the multiple measurement vector problem, which aims at recovering jointly sparse vectors from incomplete measurements. Inspired by MUSIC (MUltiple SIgnal Classification) and the greedy algorithms used in compressed sensing, we propose an empirical algorithm, called orthogonal greedy MUSIC (OG-MUSIC), for solving the problem. The proposed algorithm is a greedy algorithm, and a MUSIC procedure and an orthogonal projection operation are applied in each iteration. Since MUSIC is used in each iteration, multiple support elements may be selected per iteration; this is one of the main advantages of OG-MUSIC. The other main advantage of OG-MUSIC is the pruning technique, which is used to find the exact row support when the merged support size is larger than the sparsity level. Theoretical analysis and simulation results illustrate that OG-MUSIC has a very good recovery performance while maintaining a relatively low computational cost. © 2013 Springer Science+Business Media New York. |
Year | DOI | Venue |
|---|---|---|
2013 | 10.1007/s00034-013-9616-1 | CSSP |
Keywords | Field | DocType |
Multiple measurement vectors,Sparse recovery,Compressed sensing,Greedy algorithm,MUSIC | Mathematical optimization,Multiple signal classification,Orthographic projection,Pattern recognition,Algorithm,Greedy algorithm,Artificial intelligence,Greedy randomized adaptive search procedure,Compressed sensing,Mathematics | Journal |
Volume | Issue | ISSN |
32 | 6 | 15315878 |
Citations | PageRank | References |
0 | 0.34 | 9 |
Authors | ||
4 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Xinpeng Du | 1 | 27 | 3.90 |
| Ruihua Liang | 2 | 0 | 0.34 |
| Lizhi Cheng | 3 | 290 | 34.84 |
| Su Fang | 4 | 61 | 5.73 |