Monte Carlo Dynamic Localization Algorithm Based on Constructing Regional Filter in Underwater Acoustic Sensor Networks

How to accurately detect unknown nodes has emerged as one of the most pressing issues to be resolved in underwater acoustic sensor networks (UASNs). In this paper, a Monte Carlo dynamic localization based on constructing regional filter (RF-MCL) is proposed to address the dynamic characteristics of UASNs. In the prediction phase, based on the initial locations of the nodes, the direction condition determination is proposed to minimize the sample region and further predict the locations accurately. According to the relationship between TOA positioning error and distance, the maximum sector angle theorem is proposed and the sector sampling area is constructed based on the maximum sector angle and the maximum speed of node motion. In the filtering phase, by using ring parameters, an efficient ring region filter is constructed. The results of simulation show that compared with the existing many algorithms, the RF-MCL algorithm can achieve less localization error in lower anchor node density, under the maximum speed variation of nodes, and when the communication radius changes, and the algorithm has low complexity and good localization performance.