ABSTRACT
The past decades have witnessed rapid growth in the utilization of sensor networks consisting of a large number of sensing nodes geographically distributed in certain areas. Sensor networks have found extensive applications in various fields ranging from information collection, environmental monitoring, and industrial automation to intelligent buildings. The practical significance of sensor networks has recently led to considerable research interest in the distributed estimation or filtering problems whose aim is to extract true signals based on the information measurements collected/transmitted via sensor networks. Compared with traditional filtering algorithms in a single sensor system, the key feature of distributed filtering over sensor networks is that each sensor estimates the system state based not only on its own measurement but also on the neighboring sensors’ measurements according to the topology. So far, much effort has been made to investigate the distributed filtering problems and several effective strategies have been developed. It is worth mentioning that, up to now, the resource efficiency issue has not been adequately addressed in relation to distributed filtering problems especially for nonlinear time-varying systems, and this gives rise to the primary motivation of our current research.
