Intelligent Security System for Preventing DDoS Attacks for 6G Enabled WBSN using Improve Grey Wolf Optimization

• Published in: IEEE transactions on consumer electronics p. 1
• Main Authors: Muniyandi, Amutha Prabakar, Balusamy, Balamurugan, Dhanaraj, Rajesh Kumar, D, Sumathi, S, Nandakumar, S, Preetha K, Alroobaea, Roobaea, Paramasivam, A.
• Format: Journal Article
• Language: English
• Published: IEEE 18-06-2024
• Subjects: 6G mobile communication; 6G Technology; Consumer electronics; Denial-of-service attack; Electronic mail; Intelligent Security System; Local Search GWO; Medical Service Sector; Monitoring; Optimization; Security; Swarm Intelligent Algorithms
• ISSN: 0098-3063; 1558-4127
• DOI: 10.1109/TCE.2024.3416549

Wireless Body Sensor Network (WBSN) has gained increasing attention in health monitoring and service sector for real-time health monitoring and providing high-quality service. 6G wireless communication technologies overcome the limitations of the current 5G technology. An intelligent security system that prevents DDoS attack by applying solutions based on artificial intelligent is proposed. This method is used to identify the attackers in a heterogeneous environment by applying swarm intelligent-based Meta heuristic artificial intelligent algorithm called grey wolf optimisation (GWO). The improved GWO algorithm fine tunes user categorisation based on the service request from the WBSN devices. This approach includes KD tree (KDT) neighbour search algorithm to find the exact DDoS-attacked user electronic device by applying vertical and horizontal group division policy. The proposed method is evaluated with unsupervised dataset comprising 560 service requests, which include normal, anomaly and abnormal requests in proportions of 65%, 25% and 10%, respectively. The evaluation results show that the precision, recall, F1 Score and accuracy parameters were 92.3%, 95.2%, 93.7% and 91.5%, respectively for the predicting the abnormal electronic devices. Statistical analysis shows that the mean absolute error value is smaller than the other variants of GWO algorithms.