Multidimensional Scaling and Subnet Stitching for Enhanced Cooperative Localization
Che Zhang, Peng Han · 2025
Cooperative localization can play a significant role in GNSS-denied environments and has important application prospects. The Multidimensional Scaling algorithm is the most commonly used cooperative localization algorithm; however, it requires a complete distance matrix without missing elements, which is difficult to ensure in practical wireless ranging processes. This paper introduces a centralized processing method that divides the distance matrix with missing elements into several submatrices without missing elements. The MDS algorithm is then applied individually to each submatrix to generate the network topology of each subnet. Finally, the subnets are stitched together using the common nodes between them to achieve relative localization of the entire network. Simulation results validate the effectiveness of this method.