传感器位置误差情况下基于多维标度分析的时差定位算法  

Multidimensional Scaling Based TDOA Localization Algorithm with Sensor Location Errors

作　　者：朱国辉[1] 冯大政[1] 聂卫科 

机构地区：[1]西安电子科技大学雷达信号处理国家重点实验室,陕西西安710071 [2]西北大学信息科学与技术学院,陕西西安710127

出　　处：《电子学报》2016年第1期21-26,共6页Acta Electronica Sinica

基　　金：国家自然科学基金(No.61271293;No.61373177)

摘　　要：传统时差定位方法一般是在假设传感器位置信息准确已知的前提下进行的.然而在实际情形中,传感器位置信息往往含有随机误差,这些误差会严重影响对目标的定位精度.针对这一问题,提出了一种传感器位置误差情况下的多维标度时差定位算法.首先利用传感器位置和时差构造对称标量积矩阵,然后利用子空间理论建立关于目标位置的伪线性方程,最后通过设计加权矩阵来减少传感器位置误差对目标定位精度的影响.采用一阶小噪声扰动理论求出了目标位置估计的偏差及协方差矩阵,并通过仿真实验验证了该算法的有效性.Conventional location algorithms are based on the assumption that the sensor locations are exactly known. How-ever,in practical situations,the sensor positions generally include random errors,which can considerably reduce the source localization accuracy. To tackle this problem,a multidimensional scaling analysis based time difference of arrival（ TDOA）localization algorithm with sensor location errors is proposed. The proposed algorithm firstly constructs a symmetric matrix using the true sensor locations and TDOAs. Then a set of pseudo-linear equations with respect to the source position is formulated from the subspace theory. Finally,a weighting matrix is designed to mitigate the influence of the sensor location errors on the localization accuracy. The estimation bias and covariance matrix are derived by using first-order perturbation analysis. Simulation study validates the efficiency of the proposed algorithm.

关 键 词：无源定位 到达时间差 传感器位置误差 多维标度分析 加权最小二乘估计 

分 类 号：TN911.7[电子电信—通信与信息系统]