Summation Detector for False Data-Injection Attack in Cyber-Physical Systems.

Area of Science:

• Cyber-Physical Systems Security
• Network Intrusion Detection
• Signal Processing

Background:

• False data-injection attacks threaten cyber-physical systems (CPSs) by manipulating sensor data, leading to instability.
• Traditional detectors like the chi-squared (χ²) detector struggle against stealthy attacks that modify data subtly.

Purpose of the Study:

• To propose a novel Summation (SUM) detector for enhanced detection of false data-injection attacks in CPSs.
• To improve the resilience and security of CPSs against sophisticated cyber threats.

Main Methods:

• Development of the SUM detector, which integrates current and historical data for threat identification.
• Analysis of the SUM detector's statistical properties, showing it follows a chi-squared (χ²) distribution under normal conditions.
• Introduction of an improved false data-injection attack with a time-variable increment coefficient.

Main Results:

• The SUM detector effectively identifies both traditional and improved false data-injection attacks.
• The detector's false alarm rate can be controlled by adjusting the threshold value.
• Simulation results demonstrate the SUM detector's superior performance and effectiveness.

Conclusions:

• The proposed SUM detector offers a robust and effective solution for detecting false data-injection attacks in CPSs.
• The SUM detector provides enhanced security by leveraging historical data, outperforming existing methods.