A MDS-based localization algorithm for underwater wireless sensor network

This study is to propose a multidimensional scaling localization algorithm based on cluster structure [MDS-MAP (C, E)], and used in underwater wireless sensor networks (UWSN) node self-positioning. MDS-MAP (C, E) algorithm: First, underwater wireless sensor network is divided into more than one cluster head of the local network; Secondly, local positioning, using the Euclidean algorithm to calculate the Euclidean distance between each node and its two-hop neighbors, and then take this Euclidean distance for multidimensional scaling in each cluster, instead of the shortest distance from Dijkstra or Floyd algorithm. Simulation results show that: the localization algorithm achieves self-positioning of the whole network in the case of underwater anchor node sparse sensor nodes; the algorithm can obtain higher positioning accuracy than traditional MDS-MAP algorithm, at the same time reducing the computational overhead.