Information Fusion Estimation for spatially distributed cyber-physical systems with communication delay and bandwidth constraints

This paper studies the distributed fusion Kalman filtering problem for a class of cyber-physical systems (CPSs), where multiple sensors are arbitrarily deployed to jointly sense the state of underlying physical systems, and the sensor measurements are directly sent to the corresponding sink node. When each local estimate obtained by the sink node is transmitted to a remote information fusion center (FC), the communication between the sink node and the FC is subject to delay and finite bandwidth. A stochastic dimensionality reduction strategy is proposed to model the communication delay and bandwidth constraints, and then a recursively distributed fusion Kalman filter (DFKF) is designed from the optimal fusion criterion weighted by matrices. Since the estimation performance directly impacts the stability of control operation in CPSs, a probability-dependent sufficient condition is derived such that the mean square error of the designed DFKF is convergent. In this case, the DFKF can guarantee a satisfactory estimation performance if the selection probabilities are determined by the proposed probability-dependent condition.