Recursive state-fault filtering for stochastic nonlinear systems over sensor networks with sensor resolution and time-correlated fading channels

This paper investigates recursive resilient state estimation and fault estimation problem for a class of stochastic nonlinear systems over sensor networks under the sensor resolution and the time-correlated fading channels. The sensor resolution is regarded as a key indicator of measurement accuracy and effectively handled by the upper boundary technique to deal with the uncertainty caused by sensor resolution in this paper. The measurement output is transmitted to the remote filter through a time-correlated fading channel, where the channel coefficient is described by a certain dynamical process. The purpose of this paper is to design a resilient distributed filter to ensure that the filter error covariance has an upper bound in the presence of some fluctuations of the gain parameters. With the help of trace operations, the gain parameters of the desired distributed filter are obtained in terms of the solutions to a set of recursive equations. In addition, the exponential boundedness in the mean square of the filtering error system is analyzed. Finally, an example is given to verify the effectiveness of the proposed method.