Privacy-preserving state estimation with unreliable channels.

Area of Science:

• Information Security
• Control Systems Engineering
• Signal Processing

Background:

• Remote state estimation of dynamic systems is crucial but faces challenges from unreliable and insecure communication channels.
• Data transmission risks include message loss and eavesdropping, compromising both system state estimation and data privacy.
• Existing methods often struggle to balance privacy preservation with accurate estimation performance.

Purpose of the Study:

• To co-design a privacy-preserving encoding scheme and a filtering algorithm for secure state estimation.
• To develop a novel encoding approach that ensures information privacy and low computational cost.
• To guarantee robust estimation performance despite communication channel uncertainties.

Main Methods:

• A novel encoding scheme using weighted innovation with a public key was introduced for enhanced privacy and efficiency.
• A Minimum Mean Square Error (MMSE) estimation algorithm was designed to work with the proposed encoding mechanism.
• Performance analysis of the encoding and filtering methods was conducted.

Main Results:

• The proposed weighted innovation encoding with a public key effectively preserves information privacy.
• The developed filtering algorithm maintains reliable estimation performance under insecure and unreliable channel conditions.
• Numerical examples demonstrated the feasibility for both stable and unstable dynamic systems.

Conclusions:

• The co-designed encoding and filtering approach offers a viable solution for privacy-preserving state estimation.
• The method achieves a balance between information security, computational efficiency, and estimation accuracy.
• The approach is validated for practical applications in dynamic system monitoring and control.