Sensors Localization and Target Tracking in Underwater Environment via Belief Propagation

This paper addresses the problems of sensor localization and multi-target tracking in multi-agent networks, with primary focus on the maritime environment. Agents have access to navigation and communication measurements to reconstruct the network geometry, and at the same time exploit indirect echos of the communication signal to detect passive targets. The proposed solution merges the self-localization and tracking tasks in a single methodology based on the belief propagation algorithm. Validation on experimental measurements collected in an underwater environment proves the efficacy of the proposed solution.