Abstract
The energy-efficient tracking and precise localization of continuous objects have long been key issues in research on wireless sensor networks (WSNs). Among various techniques, significant results are reported from applying a clustering-based object tracking technique, which benefits the energy-efficient and stable network in large-scale WSNs. As of now, during the consideration of large-scale WSNs, a continuous object is tracked by using a static clustering-based approach. However, due to the restriction of global information sharing among static clusters, tracking at the boundary region is a challenging issue. This paper presents a complete tracking and localization algorithm in WSNs. Considering the limitation of static clusters, an energy-efficient incremental clustering algorithm followed by Gaussian adaptive resonance theory is proposed at the boundary region. The proposed research is allowed to learn, create, update, and retain clusters incrementally through online learning to adapt to incessant motion patterns. Finally, the Trilateration algorithm is applied for the precise localization of dynamic objects throughout the sensor network. The performance of the proposed system is evaluated through simulation results, demonstrating its energy-efficient tracking and stable network.