Energy-aware object tracking algorithm using heterogeneous wireless sensor networks

Mobile target tracking consists in detecting and locating targets in a surveillance region. Wireless Multimedia Sensor Networks (WMSN) can provide more information about the moving target than classical Wireless Sensor Networks (WSN), but their process and the transmission of multimedia data are costly which reduces the network's lifetime. In this paper, we propose a low-cost new solution for tracking a mobile target called EAOT (Energy Aware Object Tracking). It consists of a distributed cooperative algorithm that runs in heterogeneous Wireless Sensor Networks composed of both scalar and multimedia sensors. The scalar sensors (MS) are equipped with a motion detector; their role is to detect the target and then activate the camera sensors (CS) through message exchanges. We conducted simulations to evaluate the performance of the proposed mechanism in terms of tracking accuracy, number of exchanged messages, energy consumption, and spatial average deviation. The results are compared to two existing solutions: one based on simple camera sensors which always run in active mode and the other one based on the optimal camera node selection (OCNS) to track the target. The obtained results show that for the same number of nodes, EAOT consumes between 2.6% and 44% less energy than OCNS depending on the distribution of MS/CS.