A node location optimization algorithm based on mobility prediction for underwater wireless sensor networks

Because of their high flexibility and strong self-organization ability, underwater wireless sensor networks (UWSNs) are widely used in the marine field. The underwater monitoring data is of practical value thanks to UWSNs' node location technology. UWSN node location technology faces many challenges, such as node mobility, high node energy consumption and difficult to supplement node energy. This paper presents an optimization algorithm for node location in UWSNs (MPOL). Firstly, the motion prediction model of the node is established, and the kite node location is realized through the prediction and transmission of the model coefficients, so as to improve the positioning accuracy and reduce the communication consumption. Secondly, the prediction window used in the prediction stage of the located node is overlapped in the adjacent parts to eliminate the edge effect. In addition, based on the basic estimated location, the location is estimated after spatial expansion to improve the node positioning accuracy. Finally, because some nodes are beyond the location range due to mobility, this paper proposes a speed prediction method to improve the network location coverage. The algorithm effectively improves search efficiency and location accuracy while maintaining a good energy balance, according to experimental results.