Event-Based Optimal Stealthy False Data-Injection Attacks Against Remote State Estimation Systems.

Area of Science:

• Cyber-Physical Systems Security
• Control Theory
• Networked Systems

Background:

• Cyber-physical systems (CPS) face critical security challenges, with false data-injection (FDI) attacks being a significant threat.
• Degrading remote state estimation performance is a key objective for malicious attackers targeting CPS.

Purpose of the Study:

• To devise an efficient FDI attack scheme to maximize the degradation of remote state estimation performance in CPS.
• To develop an event-based stealthy FDI attack mechanism that dynamically injects false data.

Main Methods:

• An event-based FDI attack mechanism is proposed, which selectively injects false data based on real-time residuals to evade anomaly detectors.
• The state estimation error covariance is used as a metric for performance degradation.
• An optimal strategy is presented to maximize the trace of the state estimation error covariance under stealthiness constraints.

Main Results:

• The proposed event-based FDI attack mechanism effectively degrades remote state estimation performance.
• The dynamic attack strategy outperforms existing methods in maximizing performance degradation.
• Simulation results validate the superiority of the proposed method in degrading state estimation accuracy.

Conclusions:

• The developed event-based stealthy FDI attack strategy offers an efficient approach for attackers to degrade CPS performance.
• The findings highlight the vulnerabilities in remote state estimation and the need for robust security measures against sophisticated FDI attacks.