Secured dynamic event-triggered dissipative synchronization for switched neural networks subject to uncertain parameters and ADT constraint

This paper tackles the secured dissipative synchronization problem for delayed switching neural networks in the presence of random deception attacks and average dwell-time mechanism. To start with, considering that the implantation of attack signals into the controller may lead to changes in control signals, a Bernoulli distributed random variable with known statistical properties is introduced to determine whether the controller suffers from involved deception attacks. Furthermore, to enhance the usage efficiency of finite network communication bandwidth, an effective/practical dynamic event-triggered scheme is proposed, which can effectively prevent Zeno behavior. Meanwhile, this triggering scheme expands many related achievements. What is more, by resorting to Lyapunov stability principles, stochastic analytical approaches, and matrix inequality restrictions, multiple sufficient conditions can be addressed to ensure master–slave systems can achieve dissipative synchronization. By the end point, a demonstrative example is provided to confirm the validity of the suggested dynamic event-driven control approach. Moreover, for the first time, this paper also shows that the connected deception attacks have some specific influences on the amount of triggering incidents and dissipation performance index.