Secure State Estimation of Nonlinear Cyber-Physical Systems Against DoS Attacks: A Multiobserver Approach

Area of Science:

• Control Systems Engineering
• Cybersecurity
• Networked Systems

Background:

• Cyber-physical systems (CPSs) rely on accurate state estimation for control.
• Denial-of-service (DoS) attacks disrupt wireless communication, degrading estimation accuracy in CPSs.
• Existing nonlinear observers exhibit performance degradation and potential instability under DoS attacks.

Purpose of the Study:

• To develop a robust secure state estimation method for nonlinear CPSs under DoS attacks.
• To improve estimation accuracy and ensure stability during bounded-duration DoS attacks.
• To address the limitations of existing observers that become unstable during attack periods.

Main Methods:

• A novel multiobserver scheme incorporating a hold-input mechanism.
• A switched algorithm utilizing cascade observer techniques.
• Modeling CPSs as nonlinear strict-feedback systems with wireless communication.

Main Results:

• The proposed observer scheme ensures exponential convergence of the estimation error, even during DoS attacks.
• Significantly improved estimation performance compared to existing nonlinear observers.
• Demonstrated stability of the estimation error system throughout the attack duration.

Conclusions:

• The novel multiobserver and switched algorithm effectively maintain secure state estimation in CPSs under DoS attacks.
• The proposed approach enhances system resilience and reliability in the presence of cyber threats.
• Simulation results validate the theoretical findings and practical applicability of the method.