Cooperative Target Tracking In A Distributed Autonomous Sensor Network PDF Download

Wireless sensor networks (WSN) have been attracting attention over the past years owing to their capability to observe the environment, process the data, and make decisions. These features have gained importance in the context of object detection and localization, surveillance, and environmental monitoring. Specifically, there is an ongoing trend of solving these problems in a distributed manner, i.e., without the use of a central node controlling all the data generated and exchanged over the network. This project presents an algorithm for tracking multiple objects in a distributed manner using a wireless sensor network. In addition to the objects' locations, the strength of the field that they generate is considered as an unknown parameter to be estimated. The proposed algorithm is based on a diffusion cooperation scheme, which has been shown to provide good performance in distributed implementations. Our algorithm is based on explicit, non-linear tracking, and as such is not restricted to a pre-specified quantization grid in space. We allow for imperfect communication links, where information packets can be lost, and rely on an efficient adaptive combination strategy to improve robust- ness to node and link failures. Performance is analyzed in terms of the mean-squared error (MSE) as a function of the signal-to-noise ratio (SNR), in situations with different number of objects moving across the field and different probability of packet loss.