LEVERAGING SPATIOTEMPORAL PATTERNS FOR CYBER ATTACK DETECTION IN DISTRIBUTION SYSTEM

COMPUTER SCIENCE AND ENGINEERING

The increasing integration of cyber-physical systems in power distribution networks has heightened their vulnerability to sophisticated cyber attacks. This project proposes a novel framework for detecting cyber attacks in distribution systems by leveraging spatiotemporal patterns inherent in system data. By analyzing both spatial dependencies across distributed nodes and temporal trends in system behavior, the model effectively identifies anomalies indicative of malicious activities. The approach utilizes advanced machine learning algorithms, including recurrent neural networks (RNNs) and graph-based models, to capture the dynamic relationships and deviations from expected operational norms. Experimental results on simulated distribution system datasets demonstrate high accuracy in detecting a variety of cyber threats, including false data injection and command spoofing attacks. This spatiotemporal approach enhances situational awareness and supports proactive cybersecurity mechanisms for critical energy infrastructure.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.