Distributed clock synchronization and ranging in time-variant wireless networks

Clock synchronization and ranging are important topics in the field of wireless networks, where internode time measurement allows both tasks to be completed in one. Such networks are generally time-variant due to possible changes in environmental conditions and mobility of network nodes. We present a fully distributed filtering algorithm for combined clock synchronization and ranging based on message passing by belief propagation on a factor graph representation of a time-variant wireless network. The resulting message passing equations can be interpreted as a variant of Kalman filtering locally on each network node. Simulation results show that tracking estimation parameters improves estimation accuracy significantly without additional communication effort.