Adaptive joint sparse recovery algorithm based on Tabu Search

This paper introduces a novel metaheuristic methodology to address the Multiple Measurement Vectors problem using a well-known greedy search strategy. A modified version of Tabu Search algorithm is utilized to determine a precise estimation of row support and then joint sparse samples are reconstruc...

Full description

Saved in:
Bibliographic Details
Published in:Neurocomputing (Amsterdam) Vol. 224; pp. 9 - 18
Main Authors: Ghadyani, Mohsen, Shahzadi, Ali
Format: Journal Article
Language:English
Published: Elsevier B.V 08.02.2017
Subjects:
Adaptive two-step framework
Joint sparse recovery
Minimum number of samples
Multiple Measurement Vectors
Row support recovery
Tabu Search
Tabu Search
Row support recovery
Joint sparse recovery
Adaptive two-step framework
Multiple Measurement Vectors
Minimum number of samples
ISSN:0925-2312, 1872-8286
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper introduces a novel metaheuristic methodology to address the Multiple Measurement Vectors problem using a well-known greedy search strategy. A modified version of Tabu Search algorithm is utilized to determine a precise estimation of row support and then joint sparse samples are reconstructed using MMSE criterion. The proposed approach is more robust to the sparsity order variations and noise uncertainty, in comparison with the conventional MMV problem solvers. Furthermore, to avoid wastage of the sampling resources and reduce the implementation costs, a two-step joint sparse recovery framework is developed which the first step predicts and adjusts the optimum sampling rate and the second one reconstructs signal vectors applying obtained sampling rate. Numerical simulations demonstrate the superiority of proposed technique for both improving the performance and reducing the computational complexity and sampling costs.
ISSN:0925-2312
1872-8286
DOI:10.1016/j.neucom.2016.10.056