A global profile-based algorithm for energy minimization in object tracking sensor networks

Recent progress in sensing devices has made wireless sensor networks (WSNs) possible. One of their most prominent applications is object tracking. Scarce battery resources available in the sensor nodes raise the problem of energy consumption minimization in WSNs. In this paper, we exploit long term patterns in the objects’ movements to reduce the network’s energy utilization. Once the nodes have learned the objects’ movement patterns, their future location is predicted and only those nodes in the network that are expected to host the objects are woken up to detect and track them. We model the problem in an analytical way and formulate the cost of tracking the object during an application’s tracking period. We also perform extensive simulations that confirm the significant improvement brought by our proposed scheme as compared to other existing approaches.