Analysis of Hybrid Machine Learning Methods for Identifying DDoS Attacks on the SDN Control Plane
Abstract
By separating the control plane from the data plane, Software Defined Networks (SDNs) provide an attractive alternative to conventional networks. Using this function, the controller may see the whole network as a whole. Because of its essential role in the SDN ecosystem, the controller is a prime target for malicious actors. In an SDN setting, a Distributed Denial-of-Service (DDoS) assault is the most likely threat. After a Denial of Service (DDoS) assault, the legitimate user is locked out of the system for an indefinite period. In this research, we suggest using a hybrid machine learning approach to defend the controller from DDoS assaults. Our experiments also show that the hybrid model outperforms the basic models in terms of accuracy, detection rate, and false alarm rate.