Identification of the Optimum Relocalization Time in the Mobile Wireless Sensor Network Using Time-Bounded Relocalization Methodology

Contrary to the static sensor network that requires one-time localization, a mobile wireless sensor network (MWSN) requires an estimation of the optimum time to retrigger the localization of the network to accurately identify the sensor location after certain movements. However, triggering relocalization at predefined time intervals without proper consideration of the dynamic movement of sensors is insubstantial and results in poor resource management. In this paper, a new algorithm called time-bounded relocalization is proposed to identify the optimum relocalization time for the entire MWSN using the time-bounded localization method based on the analysis of the sensors' mobility pattern. In the proposed algorithm, the optimum retriggering time across the entire network can be calculated in two phases: local and global relocalizations. In the first phase, an island-based clustering method is used to estimate the local relocalization time. Next, the estimated local times are then used to decide on the optimum global relocalization time based on the statistical property of the estimated local times. For these calculations, a probabilistic model of the random waypoint (RWP) is selected. The soundness of the proposed algorithm is initially validated by deriving the probabilistic model of the optimum retriggering time, and its accuracy is checked by the Cramer–Rao lower bound (CRLB). The proposed algorithm is then extensively tested by computer simulation using practical network parameters, including the number of nodes, the size of the network, and various sizes of islands, depending on the sensor mobility, to yield the respective optimum relocalization time. The simulation results show that by using the identified optimum relocalization time, the location estimation error can be reduced by up to 32% for the RWP model, as compared with the case of using fixed relocalization time.