The location information of sensor nodes plays a vital role in the monitoring activities of wireless sensor networks. The wireless sensor network positioning algorithm based on multi-dimensional scale is very important in the multi-target positioning of wireless sensor network. Based on the analysis of the classic MDS-MAP positioning algorithm, the paper proposes the MDS-MAP(D) positioning algorithm based on Hop-Euclidean. Aiming at the disadvantages that the distributed weighted MDS positioning algorithm cannot adapt to changes in network connectivity, irregular network topology and slow convergence speed, the paper proposes an improved algorithm. The weighting mechanism and neighbor selection mechanism adopted in this paper comprehensively consider the number of 1-hop neighbors, the node’s own positioning accuracy and ranging error, and the steepest descent method is introduced to optimize its objective cost function. At the same time, the paper uses the MATLAB simulation platform to simulate and analyze the two improved algorithms proposed in terms of positioning error and topology and compare them with the original algorithm. The simulation results show that the proposed algorithm has good adaptability to irregular and large-scale networks with improved positioning accuracy.

传感器节点的位置信息在无线传感器网络的监测活动中起着至关重要的作用。基于多维尺度的无线传感器网络定位算法在无线传感器网络的多目标定位中非常重要。在分析经典MDS-MAP定位算法的基础上，提出了基于Hop-Euclidean的MDS-MAP(D)定位算法。针对分布式加权MDS定位算法不能适应网络连通性变化、网络拓扑不规则、收敛速度慢等缺点，提出一种改进算法。本文采用的加权机制和邻居选择机制综合考虑了1跳邻居数、节点自身定位精度和测距误差，并引入最速下降法来优化其目标成本函数。同时，本文利用MATLAB仿真平台对提出的两种改进算法在定位误差和拓扑结构方面进行了仿真分析，并与原算法进行了比较。仿真结果表明，该算法对不规则和大规模网络具有良好的适应性，提高了定位精度。