Towards Learning Based Detection and Mitigation of DDoS Attacks in Software Defined Networks
Hema Surendrakumar Dhadhal, Paresh P Kotak, Parvez Faruki, Atul M. Gonsai, Sajal Bhatia · 2023
The Internet of Things (IoT) has grown exponentially due to usage via smart apps. The IoT resource constraints are sometimes the reason for susceptibility. Distributed Denial-of-Service (DDoS) attacks have time and again been responsible for the Internet snip-off by exploiting the resources or saturating the bandwidth. Such attacks have turned catastrophic in recent times. Distributed Denial of Service (DDoS) attacks pose an extensive threat to networks, especially the Software-Defined Networking (SDN). The proposed paper reviews the related work that targets DDoS mitigation strategies in SDN. Based on evaluation of parameters and insights gained from the comparative analysis, the paper proposed a learning-based detection and mitigation using extreme machine learning. Our proposal is an inventive approach to detect and mitigate DDoS attacks, leveraging a semi-supervised machine learning method. We conducted a thorough evaluation of our proposed method within a controlled test-bed environment that featured an emulated network topology. The evaluation is conducted on Mininet and Ryu controllers, and relevant datasets are employed for rigorous testing and analysis. We achieved a 91.28% accuracy when compared to the outcomes of our method, better than the existing state-of-the-art.