Stability Analysis of Networked Stochastic Systems With Time Delays Under Deception Attacks by Sampled-Data Control

This article focuses on the mean-square exponential stability of networked stochastic systems with time delays (NSSTDs) under nonlinear coupling and deception attacks, employing a sampled-data control strategy. A generalized Halanay inequality for NSSTDs is proposed to investigate the stability of the closed-loop system, where multiple time delays with different bounds are considered, with an incorporation of the graph theory. By the comparison of the continuous control system and the sample-data control system, the equivalence condition on the stabilities of the two systems is studied. Meanwhile, estimates for the maximum tolerable attack probability and the corresponding largest sampling period are obtained. Moreover, a qualitative analysis of various indicators for the deception attacks and the sampling period is revealed. Following this, the theorized results are applied to linear systems with multiple time delays under nonlinear coupling, and matrix inequalities for identifying the appropriate value of the control gain are provided by the generalized Halanay inequality. To show the correctness of the results, computational simulations are conducted.