Treffer: Distributed denial-of-service attack detection for smart grid wide area measurement system: A hybrid machine learning technique

Smart grid networks face several cyber-attacks, where distributed denial-of-service (DDoS) attacks distract the grid network. The synchrophasor technique protects the wide-area measurement system (WAMS) from the complex problem and addresses different issues in a grid. The DDoS attack detection strategy is complicated due to attack complexities, vendor specifications, and communication standard protocols. Attacker target phasor measurement unit (PMU) data on the phasor data concentrator (PDC) database in WAMS. However, during the cyber-attack, the framework ensures the end application uses the normal PDC datastream. The proposed attack detection technique efficiently verifies PMU-generated data in WAMS. However, numerous machine learning algorithms are used to detect DDoS attacks, but the best detection model is still given open choices. The motivation of this study: (a) which machine learning algorithm will be suitable for DDoS attack detection and (b) what would be the accuracy of training algorithms. This study presents a machine learning-based hybrid technique that achieves 83.23% accuracy. Python compiler is used to execute the proposed model, and the result shows that the proposed detection approach efficiently improves the DDoS attack detection accuracy.