Node Adjustment Scheme of Underwater Wireless Sensor Networks Based on Motion Prediction Model

Authors:
Open Access
Publisher website
Google Scholar
Abstract
With the wide application of Underwater Wireless Sensor Networks (UWSNs) in various fields, more and more attention has been paid to deploying and adjusting network nodes. A UWSN is composed of nodes with limited mobility. Drift movement leads to the network structure’s destruction, communication performance decline, and node life-shortening. Therefore, a Node Adjustment Scheme based on Motion Prediction (NAS-MP) is proposed, which integrates the layered model of the ocean current’s uneven depth, the layered ocean current prediction model based on convolutional neural network (CNN)–transformer, the node trajectory prediction model, and the periodic depth adjustment model based on the Seagull Optimization Algorithm (SOA), to improve the network coverage and connectivity. Firstly, the error threshold of the current velocity and direction in the layer was introduced to divide the depth levels, and the regional current data model was constructed according to the measured data. Secondly, the CNN–transformer hybrid network was used to predict stratified ocean currents. Then, the prediction data of layered ocean currents was applied to the nodes’ drift model, and the nodes’ motion trajectory prediction was obtained. Finally, based on the trajectory prediction of nodes, the SOA obtained the optimal depth of nodes to optimize the coverage and connectivity of the UWSN. Experimental simulation results show that the performance of the proposed scheme is superior.
Funding Information
  • National Natural Science Foundation of China (52171337, RZ2100000416, 202310589072, ZDYF2020199)
  • Natural Science Foundation of Hainan Province (52171337, RZ2100000416, 202310589072, ZDYF2020199)
  • College Students’ Innovation and Entrepreneurship Training Project of China (52171337, RZ2100000416, 202310589072, ZDYF2020199)
  • Key Project of Hainan Province (52171337, RZ2100000416, 202310589072, ZDYF2020199)

This publication has 33 references indexed in Scilit: