A novel multiobjective optimization algorithm for sparse signal reconstruction

Sparsity and reconstruction error are two main objectives to be optimized in sparse signal reconstruction. In this paper, sparse signals are reconstructed by optimizing these two objectives simultaneously. This reconstruction method mainly consists of three steps. First, a one-dimension-dominated me...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Signal processing Jg. 167; S. 107292
Hauptverfasser: Yue, Caitong, Liang, Jing, Qu, Boyang, Han, Yuhong, Zhu, Yongsheng, Crisalle, Oscar D.
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 01.02.2020
Schlagworte:
Compress sensing
Multiobjective optimization
Particle swarm optimization (PSO)
Sparse reconstruction
Particle swarm optimization (PSO)
Sparse reconstruction
Multiobjective optimization
Compress sensing
ISSN:0165-1684, 1872-7557
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Sparsity and reconstruction error are two main objectives to be optimized in sparse signal reconstruction. In this paper, sparse signals are reconstructed by optimizing these two objectives simultaneously. This reconstruction method mainly consists of three steps. First, a one-dimension-dominated method is used to find a uniformly distributed optimal compromise solution set between these two objectives. Second, the Iterative Half Thresholding method is employed to improve the sparsity. Third, a robust selection method is proposed to choose a final solution from the solution set. The proposed method is compared with eight sparse reconstruction algorithms on twelve sparse test instances. Experimental results show that the proposed algorithm is able to reconstruct both noisy and noiseless sparse signals. In addition, the effectiveness of the proposed algorithm is demonstrated in practical application instances.
ISSN:0165-1684
1872-7557
DOI:10.1016/j.sigpro.2019.107292