Adaptive event-triggered secure synchronization control for master–slave neural networks with stochastic deception attacks

Abstract

This paper deals with adaptive event-triggered secure synchronization control for master–slave neural networks vulnerable to stochastic deception attacks and quantization. First, the advanced adaptive event-triggered scheme is offered to monitor the data transmissions, in which a new triggering error is introduced. Then, considering the adaptive event-triggered scheme, quantization, conventional deception attack and the accumulated dynamic cyber-attack, a new closed-loop synchronization control model is constructed, and two independent Bernoulli-distributed variables were served to depict the stochastic deception attacks. Moreover, by the Lyapunov–Krasovskii functional approach, an improved stability criterion is derived by using a more advanced integral inequality. Based on the criterion, a reformative controller design approach is given. Finally, the Chua’s circuit is used to validate the validity of theoretical calculations.