Distributed tracking of multiple targets with sensor networks

In this article we address the localization and tracking of multiple targets coincident in the same area. We propose a distributed method that enables us to simplify the signal processing and make the system more robust. Also we address a cluster of sensors, which are led by one of them (leader node). It is responsible to estimate the target?s location. To control the coincidence of multiple objectives in the same area, we propose an algorithm that disconnect the sensing mode and predict the target position. It permits to avoid wrong measures of the sensors and we continue with a correctly tracking. We also suggest a classification method based in the speed of the target, and it permits to reduce the uncertainty of the estimation. Simulations results show the tracking of different targets and demonstrate the reliability of these strategies.