Cooperative simultaneous localization and tracking (CoSLAT) with reduced complexity and communication

Paper in proceeding, 2013

The recently introduced framework of cooperative simultaneous localization and tracking (CoSLAT) combines Bayesian cooperative agent self-localization with distributed target tracking. The original CoSLAT algorithm suffers from high computation and communication costs because it uses a particle-based message representation. Here, we propose an advanced hybrid particle-based and parametric message passing algorithm for CoSLAT in which both costs are significantly reduced. Simulation results show that the localization/tracking performance is not affected.