Optimal actuator and sensor attack strategies against controllability and observability.

Area of Science:

• Control Systems Engineering
• Cybersecurity
• Systems Theory

Background:

• Designing optimal actuator and sensor attack strategies is crucial for understanding system vulnerabilities.
• Compromising system controllability and observability with minimal cost is a key challenge.
• System resilience against attacks is increasingly important in critical infrastructure.

Purpose of the Study:

• To develop methods for identifying minimal actuator and sensor attack sets.
• To determine the minimum number of components to attack for system uncontrollability or unobservability.
• To analyze the fragility and resilience of control systems against targeted attacks.

Main Methods:

• Investigated optimal actuator and sensor attack strategies from an attacker's perspective.
• Developed forward and reverse greedy algorithms for systems with repeated eigenvalues.
• Proposed algorithms that initiate attacks from empty and full sets to reduce computational complexity.
• Evaluated methods on a cruise-mode aircraft control system.

Main Results:

• Identified minimum-size actuator and sensor attack sets for a cruise-mode aircraft control system.
• Demonstrated that attacking selected components renders the system uncontrollable and unobservable.
• Showcased the effectiveness of the proposed algorithms in compromising system integrity.
• Achieved reduced computational complexity compared to existing greedy algorithms.

Conclusions:

• The proposed algorithms effectively identify critical actuator and sensor attack sets.
• These methods enhance the understanding of system resilience and fragility.
• The findings are valuable for securing control systems against cyber-physical attacks.