Mixed Static-Dynamic Protocol-Based Tobit Recursive Filtering for Stochastic Nonlinear Systems Against Random False Data Injection Attacks

Abstract

In this paper, the Tobit recursive filtering (TRF) issue is discussed for a class of time-varying stochastic nonlinear systems (SNSs) with censored measurements and random false data injection attacks (FDIAs) under the mixed static-dynamic protocol. The censored measurements considered are depicted by the Tobit Type I model and the phenomenon of the random FDIAs involved is governed by a set of Bernoulli random variables. Additionally, in order to reduce the communication burden and improve the data utilization efficiency, the mixed static-dynamic protocol is elaborately adopted to schedule the signal transmission, which is managed by the time-triggered and event-triggered rules to further increase the flexibility of the data scheduling. The main goal of this paper is to present a new TRF approach such that, in the presence of censored measurements, mixed static-dynamic protocol and random FDIAs, a minimized upper bound of the filtering error covariance (FEC) can be obtained. Moreover, a sufficient criterion from the theoretical analysis perspective is established to guarantee the desired uniform boundedness of the filtering error in the mean-square sense (MSS). Finally, some experiments with comparisons applicable for three-wheeled Ackerman turning model are conducted to show the applicability and advantages of newly proposed TRF scheme.