Nesnelerin interneti içeren yazılım tanımlı ağlarda gerçek zamanlı saldırı tespiti ve önlenmesi sistemi

The significant advances in wireless networks with the 5G networks have made possible a variety of new IoT use cases. 5G and beyond networks will significantly rely on network virtualization technologies such as SDN and NFV. The prevalence of IoT and the large attack surface it has created calls for SDN-based intelligent security solutions that achieve real-time, automated intrusion detection and mitigation. In this thesis, we propose a real-time intrusion detection and mitigation system for SDN, which aims to provide autonomous security in the IoT networks. The proposed approach is built upon automated flow feature extraction and classification of flows using random forest classifier at the SDN application layer. We present an SDN-specific dataset we generated for IoT and provide performance of the proposed intrusion detection model. In addition to the model performances, we provide network experiment results in the presence and absence of our proposed security mechanism. Experiment results demonstrate that the proposed security approach is promising to achieve real-time, highly accurate detection and mitigation of attacks in SDN-managed IoT networks.