Set-Membership Estimation for Complex Networks Subject to Linear and Nonlinear Bounded Attacks

Area of Science:

• Control Systems Engineering
• Network Security
• Estimation Theory

Background:

• Complex networks are vulnerable to unknown but bounded attacks in communication channels.
• Adversarial modifications to transmitted measurements pose significant challenges for accurate estimation.
• Existing estimation methods may not adequately address these sophisticated attack scenarios.

Purpose of the Study:

• To develop a novel set-membership estimation model resilient to unknown but bounded attacks.
• To establish conditions guaranteeing the existence of such estimators for linear and nonlinear attacks.
• To propose strategies for designing effective set-membership estimator gains.

Main Methods:

• Formulation of a new set-membership estimation framework.
• Derivation of sufficient conditions for estimator existence under linear and nonlinear attack functions.
• Development of two distinct strategies for designing estimator gains.
• Validation through simulation examples.

Main Results:

• A novel set-membership estimation model against unknown but bounded attacks is presented.
• Sufficient conditions are derived to ensure the existence of the estimators.
• Two effective strategies for designing the set-membership estimator gains are proposed.
• Simulation results confirm the method's effectiveness.

Conclusions:

• The proposed set-membership estimation approach effectively addresses unknown bounded attacks in complex networks.
• The derived conditions and design strategies provide a robust framework for secure estimation.
• The validated method enhances the reliability of complex network systems under adversarial conditions.