ABSTRACT
This chapter addresses the distributed target estimation problem for linear time-varying systems in Mobile Sensor Network. It discusses the system model is given and the diffusion Kalman filtering (DKF) algorithm. In, the stability of the DKF algorithm is proven under the assumption that the local Kalman filter for each agent is stable based on its local measurement information, that is, the measurements from its neighbors which include the measurement of itself according to the definition of neighbors. In the case of no local uniform observability, one choice to make the DI-DKF algorithm applicable is to build the local uniform observability for some agents in the network. A time-invariant unstable system model is considered for the ease of simulation, though the proposed algorithm is not restricted to it. The covariance intersection -DKF algorithm can be applied in the case of lacking local observability. A consensus-based information diffusion scheme is embedded when no single agent can observe the state.
