Abstract: A key obstacle to high accuracy location is the Non Line Of Sight（NLOS） transmission of signal in wireless sensor network node localization.NLOS error is the main source of location error.This paper proposes a new residual weighting location algorithm based on analysis of existing NLOS error mitigation techniques.This algorithm gradually reduces the number of range measurements from the combinations，picking out the combinations of range measurements with the minimum residual at every step and then calculating the weighted mean.In addition，it uses improved LS method to compute the temporary coordinates.Simulation results show that it has quite well location accuracy and is robust to different NLOS error.

摘要： 在无线传感器网络定位中，节点精确定位面临的一个主要问题是信号的非视距传播，非视距误差是节点定位误差的主要来源。在分析非视距误差抑制技术的基础上，提出了一种新的残差加权定位算法。算法通过逐步减少距离测量值个数的组合方式，在每一步选取残差最小的组合，再用加权平均的方法，并利用改进的最小二乘法计算节点临时坐标。仿真结果表明，算法具有较高的定位精度，在不同的非视距误差下表现稳定。