一种NLOS环境下的TDOA/AOA定位算法

TDOA/AOA Location Algorithm in NLOS Environment

在LOS环境下,Chan算法有着较好的定位精度,基于Chan算法的到达时间差/到达角(TDOA/AOA)算法比Chan算法有了进一步提高。但是在NLOS环境下,这些算法的精度都将大大下降,由于AOA的测量值有较大误差,TDOA/AOA方法的精度甚至低于Chan算法。并且这些算法的主要缺点是在第一次加权最小二乘法(WLS)中把移动台的横坐标、纵坐标与移动台到服务基站的距离作为三个相互独立的变量,忽略了三者之间的相关性,因此要进行第二次WLS才能得到定位结果,且最终的解为二值根。对误差的均值和方差进行了估计,修正了TDOA与AOA测量值,用Kalman滤波算法对AOA的值进行了估计,利用移动台坐标与AOA之间的关系将三个变量简化为一个,只需一次WLS即可求得唯一解,减少了计算量,消除了根的模糊性。仿真结果表明,该方法简单,计算量小,有较高的定位精度和较好的稳健性,性能优于Chan算法和基于Chan算法的TDOA/AOA算法。

In LOS environment,Chan＇s algorithm has goodish location accuracy and TDOA/AOA algorithm hased on Chan＇s has better accuracy than it.But in NLOS environment,location accuracy of these algorithms degrade greatly.Because the measurement of AOA has large error,the location accuracy of TDOA/AOA method is even worse than that of Chan＇s.Tbese algorithms assume the abscissa,the ordinate of the MS and the distance between the MS and the servering BS to be three independent variables neglecting their correlation in the first step of the Weighted least Square （WLS） method.So the WLS method must be performed twice with each root has two values.This paper estimates the mean and variance of error and corrects the measurements of TDOA and AOA.Then it estimates the AOA by using Kahnan filter and simplifies the three variables into one according to the relationship among the coordinates of the MS and AOA.The proposed algorithm needs less calculation amount because it uses the WLS method only once.So it can get the only root and eliminate the ambiguity.The result of simulation proves its simplification,less calculating amount, better location accuracy and robustness.The performance of this algorithm is better than that of Chan＇s and the TDOA/ AOA algorithm based on it.