Design a Robust DDoS Attack Detection and Mitigation Scheme in SDN-Edge-IoT by Leveraging Machine Learning

The exponential growth of the Internet of Things (IoT) has significantly improved various domains, including healthcare, transportation, and smart cities. However, this rapid expansion introduces critical security vulnerabilities, with Distributed Denial of Service (DDoS) attacks emerging as a predominant threat. Traditional DDoS detection mechanisms in Software Defined Networking (SDN)-enabled IoT networks often struggle with limited detection accuracy and high latency due to centralized control plane processing. This project proposes a robust DDoS attack detection and mitigation framework in SDN-Edge-IoT environments by leveraging machine learning techniques. The system incorporates Long Short-Term Memory (LSTM) and Decision Tree algorithms, enhanced through hyper-parameter tuning and k-fold cross-validation, to ensure high detection accuracy. To address latency concerns, the trained models are deployed at the edge layer, enabling real-time anomaly detection. Upon detecting suspicious traffic, the SDN controller dynamically enforces mitigation rules to protect the network. The proposed scheme demonstrates improved efficiency in early DDoS detection and rapid response, ensuring a secure and resilient SDN-IoT infrastructure.

Keywords: DDoS Attack Detection, Software Defined Networking (SDN), Internet of Things (IoT), Edge Computing, Machine Learning, LSTM, Decision Tree, Hyperparameter Tuning, Cross-Validation, Network Security, SDN Controller