AI-Enabled DDoS Detection and Mitigation in the Software Defined Network

Traditional networks typically employ a decentralized control plane, while Software Defined Networking (SDN), a dynamic networking space, adopts a centralized control system. SDN has many benefits, including precise control, optimized routing, and efficient resource management, making it well-suited for data-centric networks. Despite these benefits, SDNs are susceptible to traffic attack, and the development of robust detection and mitigation strategies. Addressing the challenge of attack with Distributed Denial of Service (DDoS) in SDN environment. This research proposes an OpenFlow-based detection approach for mitigation within the control plane. And employs a specific SDN DDoS attack dataset that incorporates unique features. When it comes to DDoS attack detection, deep learning and machine learning models are used for classifying these attacks. The outcomes reveal that a Neural Network demonstrates good performance among the algorithm used on this dataset, achieving an accuracy of 99.4%. Subsequently, the controller takes action blocking the malicious host, employing the addition of blocking rules to the switches.