A Framework for Detecting and Mitigating DDoS Attacks in Software-Defined IoT Networks

Abstract

The integration of Internet of Things (IoT) technology and Software-Defined Networking (SDN) into Software-Defined IoT (SD-IoT) creates new opportunities while introducing significant security challenges. An emerging area of research is the detection and mitigation of DDoS attack on SD-IoT networks. Many existing DDoS intrusion detection methods rely on outdated datasets. This thesis presents a framework for DDoS detection and mitigation in SD-IoT. A novel approach is introduced for extracting informative features from network traffic and generating datasets. In the detection phase, generated datasets are used to train machine learning models for DDoS detection. For mitigation, a strategy combining micro-segmentation with Attribute-Based Access Control (ABAC) is proposed, enabling effective attack containment and establishing a robust defense-in-depth framework. The results demonstrate the importance of the extracted features in training machine learning models for DDoS detection. A use case further illustrates the efficiency of the proposed micro-segmentation method in mitigating DDoS attacks.