Abstract: Two non-metric multidimensional scaling based node localization algorithms are proposed for bestrewn wireless sensor networks，NMDS-MAP and NMDS-MAP（P）.They measure the pairwise distances between nodes with TDOA or other techniques，make use of non-metric multidimensional scaling technique to estimate the nodes location.The former algorithm is centralized，and the later distributed.Simulation experimentation demonstrates that the algorithms are more accurate and robust in positioning sensors.

摘要： 针对撒布型无线传感器网络提出了基于非度量多维标度的NMDS-MAP算法及NMDS-MAP（P）算法，两种方法采用TDOA等测距技术测量节点间距，利用非度量多维标度技术对未知节点进行定位，前者是集中式算法，后者是分布式算法。理论分析与仿真实验表明，两种算法具有较高的定位精度与健壮性。