Cooperative Simultaneous Localization and Tracking in Mobile Agent Networks

We introduce a framework and methodology of cooperative simultaneous localization and tracking (CoSLAT) in decentralized mobile agent networks. CoSLAT provides a consistent combination of cooperative self-localization (CSL) and distributed target tracking (DTT). Multiple mobile targets and mobile agents are tracked using pairwise measurements between agents and targets and between agents. We propose a distributed CoSLAT algorithm that combines particle-based belief propagation with the likelihood consensus scheme and performs a bidirectional probabilistic information transfer between CSL and DTT. Simulation results demonstrate significant improvements in both self-localization and target tracking performance compared to separate CSL and DTT.