Event-triggered resilient filtering with measurement quantization and random sensor failures: Monotonicity and convergence

Abstract This paper is concerned with the remote state estimation problem for a class of discrete-time stochastic systems. An event-triggered scheme is exploited to regulate the sensor-to-estimator communication in order to preserve limited network resources. A situation is considered where the sensors are susceptible to possible failures and the signals are quantized before entering the network. Furthermore, the resilience issue for the filter design is taken into account in order to accommodate the possible gain variations in the course of filter implementation. In the simultaneous presence of measurement quantizations, sensor failures and gain variations, an event-triggered filter is designed to minimize certain upper bound of the covariance of the estimation error in terms of the solution to Riccati-like difference equations. Further analysis demonstrates the monotonicity of the minimized upper bound with respect to the value of thresholds. Subsequently, a sufficient condition is also established for the convergence of the steady-state filter. A numerical example is presented to verify the effectiveness of the proposed filtering algorithm.