Distributed Denial of Service Attack Detection in SDN using Machine Learning

Abstract

SDN, or software-defined networking, is an approach to networking in which network traffic is controlled by software controllers or APIs instead of traditional networking hardware. An efficient and powerful platform is one benefit of dividing the control and data planes. Software-Defined Networking (SDN) advancements are accelerating as new network architectures arise and face grave challenges such as Distributed Denial of Service (DDoS) attacks. A common malicious assault called a DDoS attempt floods the affected user’s system with Internet traffic in an effort to disrupt the regular flow of traffic to the server, network, or service. Models based on machine learning were used in this project to identify Distributed Denial of Service (DDoS) attacks in SDNs (SDN). A new dataset was produced using feature selection techniques to make models simpler, make it easier to interpret them, and save training time. Research in this area includes trying out different machine learning models and figuring out how to implement them into the DDoS detection system. Several categorization methods popular in machine learning-based DDoS attack detection in SDN were evaluated using the CIC-DDoS 2019 dataset. Finally, both the potential benefits and drawbacks of using machine learning-based detection algorithms in SDN are discussed.