A distributed particle-based belief propagation algorithm for cooperative simultaneous localization and synchronization

Abstract

We present a factor graph framework and a particle-based belief propagation algorithm for distributed cooperative simultaneous localization and synchronization (CoSLAS) in decentralized sensor networks. The proposed algorithm jointly estimates the locations and clock parameters of the network nodes in a fully decentralized fashion. This estimation is based on time measurements and communications only between neighboring nodes, and makes only minimal assumptions about the network topology. A significant reduction of computational complexity can be achieved by a novel particle-based scheme for message multiplication. Simulation results demonstrate the good performance of the proposed CoSLAS algorithm.