Multi-Target Localization Based on Sparse Bayesian Learning in Wireless Sensor Networks

Bo XUE  Linghua ZHANG  Yang YU  

IEICE TRANSACTIONS on Communications   Vol.E99-B    No.5    pp.1093-1100
Publication Date: 2016/05/01
Online ISSN: 1745-1345
DOI: 10.1587/transcom.2015EBP3250
Type of Manuscript: PAPER
Category: Network
multi-target localization,  sparse Bayesian learning,  compressed sensing,  wireless sensor networks,  

Full Text: PDF(982.1KB)>>
Buy this Article

Because accurate position information plays an important role in wireless sensor networks (WSNs), target localization has attracted considerable attention in recent years. In this paper, based on target spatial domain discretion, the target localization problem is formulated as a sparsity-seeking problem that can be solved by the compressed sensing (CS) technique. To satisfy the robust recovery condition called restricted isometry property (RIP) for CS theory requirement, an orthogonalization preprocessing method named LU (lower triangular matrix, unitary matrix) decomposition is utilized to ensure the observation matrix obeys the RIP. In addition, from the viewpoint of the positioning systems, taking advantage of the joint posterior distribution of model parameters that approximate the sparse prior knowledge of target, the sparse Bayesian learning (SBL) approach is utilized to improve the positioning performance. Simulation results illustrate that the proposed algorithm has higher positioning accuracy in multi-target scenarios than existing algorithms.