Abstract
Sparse signal reconstruction algorithms in compressive sensing mainly focus on greedy algorithms, which are short-sighted. In this letter, an optimization-oriented algorithm using global optimization method is proposed to reconstruct the sparse signal. First, a pre-selection is made to estimate the most likely support of the sparse signal. Second, a backtracking strategy is introduced to avoid the over-fitting problem. Last, an optimization-oriented search strategy is designed to refine the pre-selected support toward the best estimate of the signal support. Numerical results demonstrate that the proposed algorithm performs better for sparse signal reconstruction with moderate computational complexity compared with the existing algorithms.
Original language | English |
---|---|
Article number | 8633941 |
Pages (from-to) | 515-519 |
Number of pages | 5 |
Journal | IEEE Signal Processing Letters |
Volume | 26 |
Issue number | 3 |
DOIs | |
State | Published - Mar 2019 |
Externally published | Yes |
Bibliographical note
Publisher Copyright:© 2019 IEEE.
Keywords
- Compressive sensing
- global optimization
- greedy algorithms
- optimization-oriented algorithm