Resilient discontinuous event-triggering control for exponential stabilization of memristive neural networks under denial-of-service attacks

This paper studies the exponential stabilization issue of memristive neural networks (MNNs) in the presence of denial-of-service (DoS) attacks by using an event-triggering scheme. Unlike the existing event-triggering strategies, not only does the devised resilient discontinuous event-triggering (RDET) scheme avoid the Zeno phenomenon and reduce network communication resource utilization, but can it effectively deal with the non-periodic DoS attacks. In the joint framework of RDET control and DoS attacks, a closed-loop MNNs system is established. Then, to address the attacks in different scenarios, two different-interval-dependent functionals are established. The continuity of functionals improves the anti-attack rate compared with the previous work. Moreover, by using a combination of the convex combination and the estimation techniques, the exponential stabilization results are deduced and a secure controller associated with event-triggering parameters are co-designed. Finally, simulations are carried out to demonstrate the effectiveness of the derived stabilization results and the practical advantages of the proposed RDET scheme subject to DoS attacks.