Intermittent Adaptive Resilient Secure Control for Nonlinear CPSs Under DoS Attacks in Sensor Networks

This paper investigates the problem of adaptive prescribed-time intermittent resilient tracking control for a class of nonlinear cyber-physical systems (CPSs) under full-state denial-of-service (DoS) attacks with malicious intent. Since the disturbance of DoS attacks causes the discontinuity of the system states, which leads to its non-differentiability, it certainly increases the difficulty of the subsequent controller design. To overcome this obstacle, we reconstruct the system states by designing a set of adaptive high-gain observers, which generate differentiable state estimation signals. Subsequently, a time-varying constraint function is then integrated into the backstepping framework to construct a novel adaptive resilient controller with prescribed-time performance. This approach effectively converts the prescribed-time tracking control problem into a constraint problem for the tracking error. Meanwhile, a dynamic-threshold event-triggered mechanism (ETM) is proposed to minimize redundant signal transmissions and enhance the efficiency of network resource usage. Moreover, by using the Lyapunov stability theory, the proposed security control strategy ensures asymptotic convergence of the tracking error to zero, even in the absence of prior information regarding the frequency and duration of DoS attacks. Finally, the validity and practicality of the developed scheme are verified through simulation results on a one-link manipulator system.