A Resilient Data-Driven Controller Against DoS Attacks

This paper is concerned with the stabilization problem of linear time-invariant systems under Denial-of-Service (DoS) attacks. Meanwhile, system matrices are unknown, and only some input-output trajectories collected from off-line experiments are accessible. Generally, in the model-based case, it has been proven that systems equipped with a predictor-based state feedback controller achieve resilience against DoS attacks. However, this is difficult or even impossible to be implemented in the absence of a system model, thus leading to resilient control challenging in the data-based case. To maintain resilience against DoS attacks in the data-based case, a data-driven model predictive control (MPC) scheme is proposed such that future system input-output trajectories can be obtained by solving a data-dependent optimal problem. Leveraging this scheme, a data-driven resilient control strategy is developed such that the system achieves the same level of resilience of the DoS attacks as the model-based case. Finally, a numerical example is given to validate the effectiveness of the proposed method.