Stochastic Watermarking Protection Strategy for CPSs Under Linear Stealthy Attacks

Area of Science:

• Cybersecurity
• Control Systems Engineering
• Game Theory

Background:

• Cyber-physical systems (CPSs) are vulnerable to sophisticated linear stealthy attacks.
• Deterministic watermarking is insufficient against attackers who can decipher watermarking signals.
• Stochastic watermarking offers a more robust defense mechanism for CPS security.

Purpose of the Study:

• To develop and analyze a stochastic watermarking protection strategy for CPSs.
• To model the adversarial interaction between attackers and defenders using game theory.
• To determine optimal watermarking parameters for enhanced CPS security.

Main Methods:

• Formulation of a zero-sum game to represent the attacker-defender interaction.
• Analysis of equilibrium strategies based on attack/defense probabilities and watermarking parameters.
• Derivation of criteria for selecting watermarking parameters to minimize attacker payoff.

Main Results:

• The equilibrium strategy is dependent on maximum attack/defense probabilities and time-varying watermarking parameters.
• Criteria for selecting watermarking parameters are provided to reduce the attacker's payoff.
• The proposed defense strategy is effective when attack probability is known.

Conclusions:

• Stochastic watermarking provides a viable defense against stealthy attacks in CPSs.
• Game theory is a powerful tool for analyzing and optimizing security strategies in CPSs.
• The findings offer practical guidance for implementing robust CPS security measures.