Abstract
Cooperative localisation (CL) is an effective way to enhance the position estimate by cooperation among agents, especially under insufficient signal source conditions. In centralised CL (C-CL), all the information is transmitted to a data centre to find a globally optimal solution for all agents. However, the centralised network is not scalable since the collection of all the measurements requires higher communication overhead. In distributed CL (D-CL), each agent iteratively estimates its position by exchanging this estimate with its neighbours, which require proper initial positions. In this study, cluster-based networks are adopted to solve the scalability issue of C-CL and the initial position issue of D-CL. Analysis of performance limit for D-CL is derived based on Cramér–Rao lower bound. The theoretical foundation supports the utilisation of the proposed cluster-based CL (CbCL) to decouple C-CL problem into intra-cluster CL problem for a reliable initial position first. After the proposed N-anchors cluster formation, the clusterhead estimates the cluster members’ initial positions based on intra-cluster measurements. The cluster gateway further iteratively refines its position estimate with other cluster gateways based on inter-cluster measurements. Simulation results validate that the CbCL achieves a tradeoff between the communication overhead and the localisation performance for a scalable implementation.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call