Software-Defined Networking (SDN) presents a novel approach to network architecture, characterized by the separation of control and data planes, enabling centralized management and agile resource allocation. However, this newfound flexibility also exposes networks to significant security vulnerabilities, particularly concerning Denial-of-Service (DoS) attacks, which disrupt service availability by overwhelming network resources. This study explores the application of machine learning techniques to detect and mitigate DoS attacks in SDN environments. Utilizing the InSDN dataset, a benchmark specifically designed for SDN-based intrusion detection, two widely used machine learning models—Support Vector Machine (SVM) and Random Forest (RF)—were trained to differentiate between normal and attack traffic. The experimental results indicate that the Random Forest model outperforms SVM, achieving a remarkable accuracy of 98%, while SVM follows closely with 97%. In order to resolve class imbalances present in the dataset, the research evaluates the models under other situations, such as under and over-sampling. The results demonstrate the Random Forest model's better performance across a range of data handling techniques, highlighting its resilience and adaptability. This research contributes to the growing body of research on using machine learning to improve SDN's security posture and makes a compelling argument for incorporating these advanced algorithms into existing network security frameworks. This study highlights the significance of choosing suitable models and techniques based on particular data distributions and security requirements by clarifying the advantages and disadvantages of both models. This helps practitioners strengthen SDN infrastructures against changing cyber threats.

Authors: I.A. Oshin, O.A. Odeniyi

Published in: International Conference for Internet Technology and Secured Transactions (ICITST-2024)

• Date of Conference: 4-6 November 2024
• DOI: 10.20533/ICITST.2024.0006
• ISBN: 978-1-913572-76-1
• Conference Location: St Anne’s College, Oxford University, UK