Resilient event-triggered stabilization for uncertain Markovian jump delayed neural networks under time-window-based aperiodic DoS attacks

Abstract

In this paper, the problem of robustly stable in mean square of uncertain Markovian jump neural networks (UMJNNs) with time-varying delays under time-window-based aperiodic denial-of-service (DoS) attacks is investigated. First, a new class of portrayal is proposed for DoS attacks, that is, fixed time-window-based non-periodic DoS attacks. In addition, a resilient event-triggered communication scheme (RETCS) is designed between sensors and controllers to reduce “unnecessary” waste of network resources under the proposed non-periodic DoS attack. Then, a new model of UMJNNs with time-varying delays considering non-periodic DoS attacks is developed on this basis. Second, a new single Lyapunov function is constructed in this paper for non-periodic DoS attacks. In addition, the stability criteria of UMJNNs with time-varying delays are obtained based on Lyapunov stability theory and the linear matrix inequality technique. Then, the criterion for co-designing the trigger parameters of RETCS and the gain matrix of the controller is proposed. Finally, the validity of the obtained result is illustrated by two examples.