ENHANCING SDN SECURITY: A COMPREHENSIVE FRAMEWORK FOR MITIGATING DDOS ATTACKS IN SOFTWARE-DEFINED NETWORKS

Abstract

This study presents a comprehensive framework for mitigating Distributed Denial of Service (DDoS) attacks within Software-Defined Networks (SDN), focusing on enhancing security and network performance. The framework integrates flow-based anomaly detection, machine learning models, and decentralized traffic routing to dynamically identify and mitigate DDoS attacks in real-time. The results indicate that the proposed framework achieved high detection accuracy for volumetric and protocol attacks (98.7% and 95.4%, respectively), while application-layer attacks, though more challenging, were detected with an accuracy of 93.9%. When applied to volumetric DDoS attacks the framework delivered fewer attack response times at 95.6 ms compared to traditional controls such as rate limitation and blacklisting that produced longer response periods.  The system's operation maintained low latency coupled with high throughput to ensure network stability under intense attacks.  Machine learning models specifically neural networks delivered accuracy to the framework by means of their 2.1% low false positive rate.  The proposed architectural solution proved that it could expand its capabilities without overburdening the SDN controller resources.  The system distributed DDoS defence responsibilities throughout multiple network points for attack defense functionalities and real-time traffic rerouting which preserved continuous network access.  This investigation demonstrates how the framework operates with scale and versatility across SDN systems at large so it presents effective DDoS protection solutions for current network frameworks