Hybrid Machine Learning Approaches For Classification DDoS Attack

The rapid advancements in technology and the adoption of 5G have significantly accelerated The expansion of the Internet of Things (IoT) has resulted in a significant rise in the number of connected devices. This surge has not only expanded network capabilities but also increased the potential attack surface, making networks more vulnerable to Distributed Denial of Service (DDoS) attacks. Addressing this critical issue, We introduce an innovative hybrid machine learning approach for DDoS attack detection. Our method integrates Radial Basis Function Networks (RBFNs) with Support Vector Machines (SVMs), optimized through the K-means clustering algorithm, to enhance detection accuracy and efficiency. We conducted experiments using two datasets: CICDDoS2019 and CICIDS2017. Our approach achieved impressive results, with the CICDDoS2019 dataset yielding an precision of 99.42%, accuracy of 99.68%, F1 score of 99.55%, and an recall of 99.68%. Similarly, alongside the CICIDS2017 dataset, we achieved an accuracy of 98.86%, precision of 98.17%, recall of 99.86%, and an F1 score of 99.01 %. These results demonstrate The efficiency of our proposed method in accurately detecting DDoS attacks, thereby contributing to the robustness and security of IoT networks.