A predictive localization algorithm for underwater wireless sensor networks based on improved backtracking search optimization and gray wolf optimizer

The gradual expansion of the earth's population and the increasing shortage of land resources promote people's exploration of marine resources undoubtedly. Underwater wireless sensor networks (UWSNs) technology has been developed. Its applications involve military field, civil resource detection, scientific research and experiment, etc. The implementation of related applications need to rely on UWSNs node localization, which is one of the key technologies of UWSNs. Accurate node location information is the premise of all applications. Aiming at the low location accuracy and location coverage of existing localization algorithms, a predictive localization algorithm (MGP) based on modified backtracking search optimization (MBSA) and gray wolf optimizer (GWO) for UWSNs is designed in the paper. The whole localization process includes two parts mainly: first floor nodes localization and secondary floor nodes localization. Firstly, MBSA is introduced into the first layer to improve the location accuracy and convergence speed of nodes; Secondly, GWO algorithm is introduced into the secondary layer localization. GWO algorithm optimizes the location information with low estimation accuracy. Finally, the simulation shows that MGP has better localization results than SLMP, MCL-MP and MP-PSO.