دراسة تحليلية لكشف DDoS في البيئات السلكية واللاسلكية لشبكات SDN باستخدام خوارزميات التعلم الآلي

Abstract

The effective detection of Distributed Denial of Service (DDoS) attacks has become an urgent necessity for the security of modern networks, especially in Software-Defined Networking (SDN) environments. In this article, we presented an evaluation of the effectiveness of Machine Learning (ML) algorithms in identifying and classifying these attacks within SDN environments, with a particular focus on comparing the performance of these algorithms between wired and wireless networks. To achieve this, we used two datasets: the first for a wired SDN network and the second for a wireless SDN network. Both datasets include natural traffic features and specific traffic features of DDoS attacks.

We applied five main machine learning algorithms: Decision Tree (DT), K-Nearest Neighbors (KNN), Logistic Regression (LR), Random Forest (RF), and Gaussian Naive Bayes (GNB). We evaluated the performance of these algorithms using metrics such as Accuracy, Recall, Precision, and F1-Score, in addition to training time. The results showed that the RF and DT algorithms achieved the highest levels of accuracy in both environments, with RF outperforming in overall performance. Meanwhile, KNN emerged as the fastest algorithm in the wired environment and showed a significant improvement in accuracy and speed in the wireless environment.