Target Localization Algorithm of Wireless Sensor Network Based on Fast Orthogonal Matching Pursuit in

基于快速正交匹配追踪的无线传感网中目标定位算法

School of Marine Science and Technology, Northwestern Polytechnical University, Xi'an 710072, China

Received: 20 January 2019

Abstract

Efficient localization of multiple targets is one of the basic technical problems in wireless sensor networks (WSN). The traditional sparse representation method based on greedy class is not efficient in multi-target positioning. Aiming at this problem, a multi-target localization algorithm based on QR decomposition for fast orthogonal matching pursuit is proposed. The algorithm meshes the wireless sensor coverage area to design an over-complete dictionary, which transforms the multi-target localization problem into a sparse signal recovery problem. The method utilizes the sensor to receive the sparse characteristics of the target signal strength, and then uses fast orthogonal matching pursuit to recover the measured values, thereby localization the target by sparsity. Through the QR decomposition of the column full rank matrix, the recursive form is used to invert the sub-dictionary matrix, thus avoiding the direct inversion of the matrix in the traditional method, so that the computational complexity is greatly reduced. The simulation results show that compared with the traditional orthogonal matching pursuit compressed sensing reconstruction method, this method does not lose the localization accuracy and improves the computational efficiency.

摘要

高效准确的多目标定位是无线传感器网的基本任务之一。传统基于贪婪类的稀疏表示方法在多目标定位中计算效率不高。针对该问题，提出一种基于QR分解的快速正交匹配追踪的多目标定位算法。该算法对无线传感器覆盖区域进行网格划分来设计过完备字典，从而将多目标定位问题转化为稀疏信号恢复问题。该方法利用了传感器接收目标信号强度的稀疏特性，然后使用快速正交匹配追踪来恢复测量值，进而通过稀疏性来定位目标。通过列满秩矩阵的QR分解思想，利用递归形式来对子字典矩阵求逆，避免了传统方法中对该矩阵的直接求逆，使得运算量大为降低。仿真结果表明，与传统的正交匹配追踪压缩感知重构方法相比，该方法不损失定位精度，提高了运算效率。