Toward the Integration of Cyber and Physical Security Monitoring Systems for Critical Infrastructures

Area of Science:

• Engineering
• Computer Science
• Cybersecurity

Background:

• Critical Infrastructures (CIs) are vulnerable to physical and cyber threats, leading to significant disruptions.
• Existing security measures often analyze physical and cyber domains separately, potentially missing complex threats.

Purpose of the Study:

• To develop and evaluate a novel approach for enhancing the security of critical infrastructures.
• To demonstrate the effectiveness of correlating physical and cybersecurity data for anomaly detection.

Main Methods:

• Proposed a logical architecture integrating data from physical and cybersecurity systems.
• Developed a Machine Learning (ML)-based anomaly detection model.
• Correlated data from both physical and cyber domains to identify anomalies.

Main Results:

• The ML-based approach successfully detected anomalies by correlating cross-domain data.
• The integrated approach identified abnormal situations missed by single-domain analyses.
• Evaluated performance using various ML algorithms, showing encouraging results.

Conclusions:

• Jointly analyzing physical and cyber data significantly improves anomaly detection in critical infrastructures.
• The proposed architecture offers a more robust security solution for power plant scenarios.
• Cross-domain correlation is crucial for comprehensive CI protection against sophisticated threats.