基于到达时间差的两步最小二乘定位算法

Two-step least square localization algorithm using time difference of arrival measurement

为了提高单目标定位的精度,以到达时间差为物理量,提出了一种两步最小二乘定位算法(TSLS).TSLS首先基于普通线性最小二乘估计模型,采用线性纠正最小二乘定位算法(LCLS)获得目标节点位置的估计值;然后计算出目标节点到达不同基站的距离差,并用该距离差来近似实际距离差;最后基于约束总体最小二乘估计模型,利用约束总体最小二乘定位算法(CTLS)对目标节点位置进行二次估计.由于TSLS本质上是通过减少CTLS的矩阵误差来提高CTLS的性能,因此可以看成是一种增强型CTLS算法.仿真分析表明,TSLS算法的总体性能要优于CTLS算法和LCLS算法.

In order to improve the localization precision for a single object,a two-step least square localization algorithm （TSLS）is proposed based on the time difference of arrival.The object loca-tion is first estimated by the line correct least square algorithm （LCLS）based on the common least square estimation model.Then the distance difference from the object to different base stations is computed based on the estimated location.Finally,the location is estimated by the constrained total least square algorithm （CTLS）based on the constrained total least square model,in which the dis-tance difference is approximated not by the measurement but by the calculation.Essentially,TSLS improves the performance of CTLS by reducing the errors in the matrix of CTLS,so it can be seen as an enhanced CTLS algorithm.Simulation results show that the performance of the TSLS is better than that of the CTLS and LCLS.