RIASSUNTO
In underwater sensor localization scenes, beacon nodes (BNs) with known coordinates are deployed to localize unknown nodes (UNs). However, the BNs are likely to drift due to undercurrents, geological changes, and marine biological activities, which will lead to localization failure. This paper proposes underwater localization evaluation scheme (ULES) to improve localization accuracy under drift scenes. Reliability of BN is evaluated based on underwater drift, underwater environment, and underwater acoustic channel using an analytic hierarchy process and a grey correlation method. For each UN, four BNs in its communication range with high reliability are selected for underwater localization. This process can iterate for several times, and newly localized nodes act as temporary BNs in the next iteration. Simulations are performed for different localization schemes under different BN deployment schemes and scenarios. Results show that about a quarter more UNs obtain accurate localization comparing with the conventional schemes under beacon node drift scenes, which fully shows that ULES reduces average localization error under BN drift scenes.