Multi-stage mobile beacons-assisted localization scheme for large scale underwater sensor network

Abstract

Localization is one of the most important issues in Underwater Wireless Sensor Networks (UWSN) as sensor nodes are considerably difficult to be deployed at determined locations. Beacon nodes static or mobile are responsible to disseminate their location information to unknown nodes in order to determine the location of each deployed underwater. The key challenge is to provide a scalable localization scheme to cover all sensor node with sufficient beacon message, while reducing localization error in estimating node position and decreasing localization time. Mobile-beacon assisted localization is the best option for large-scale network due to its high accuracy and costeffectiveness compared to static beacons. However, the effect of water current causes a non-uniform deployment of underwater sensor and influences the variable speed of sound in acoustic signal, and thus making the existing mobile-beacon localization schemes requires high communication cost and generates low localization accuracy. Therefore, to overcome this issue, this study proposed the improved mobile beacon assisted localization scheme. Firstly, Adaptive Underwater Cluster-based (AUC) algorithm based on communication range control was suggested for underwater sensor network to improve localization coverage and reduce the estimation error. Secondly, Efficient Dynamic Mobile beacon Path Planning (EDMPP) algorithm was put forward to improve the placement of virtual beacon point and reduce the communication overhead. Thirdly to maintain the localization coverage as well as reduce the communication overhead between mobile beacon and sensor node in large-scale UWSN Hybrid Mobile beacon Path Planning (HybMPP) algorithm was also proposed. Result on non-uniform deployment for UWSN simulation indicated that AUC improve 23% of localization accuracy compared to MDS-MAP(p) and 52% compared to MDSMAP. In beacon message transmission, EDMPP reduced communication overhead by 38.42% in comparison to 3D Hilbert; and 37.08% when compared to 3D SCAN. Moreover, EDMPP decreased localization time by 47.36% compared to 3D Hilbert and 10.48% compared to 3D SCAN. Subsequently, as compared to EDMPP in a deep sea environment, HybMPP lessened communication overhead by 56.27% for sensor node discovery. This illustrates that the suggested approach outperformed the existing algorithm in the localization process.