Attack Detection and Active Attack Defense for Cyber--Physical Systems via Zonotopic Observer and Reachability...

Area of Science:

• Cyber-Physical Systems Security
• Control Theory
• Fault Detection and Diagnosis

Background:

• Cyber-physical systems (CPS) face increasing threats from actuator and sensor attacks.
• Unknown but bounded (UBB) disturbances and noise complicate attack detection and defense.
• Existing methods often struggle to simultaneously address state estimation and attack detection.

Purpose of the Study:

• To develop robust attack detection and active defense strategies for discrete-time linear CPS.
• To simultaneously estimate system states and actuator attacks under uncertainty.
• To mitigate the impact of sensor and actuator attacks on system performance.

Main Methods:

• A novel zonotopic observer incorporating augmentation and H-infinity techniques for state and actuator attack estimation.
• Set-valued and residual-based abnormal detectors utilizing reachability analysis.
• An active attack defense mechanism based on state reachable sets and H-infinity control.

Main Results:

• Simultaneous estimation of system states and actuator attacks.
• Effective detection of both actuator and sensor attacks.
• An active defense strategy that preserves system performance in the absence of attacks.

Conclusions:

• The proposed methods offer a superior approach to attack detection and defense in CPS.
• The techniques are validated through numerical simulations and an autonomous aircraft system.
• This work advances the resilience of cyber-physical systems against sophisticated attacks.