Distributed Kalman Filtering for Noisy Consensus Networks with Time Delays

This work focuses on consensus networks consisting of a group of mobile agents in the presence of noise and time delays. Due to noise, exact consensus is never reached. Hence, we propose fully distributed and local Kalman filters that filter only a subset of all system states while maintaining a filter quality close to a global Kalman filter. Feasibility is guaranteed by communicating measurements through the network, introducing time delays. The overall system stability is analyzed. Simulations demonstrate the performance of the proposed distributed filters.