SDN-Based DDOS Attack Identification Using Random Forest Classification

Software-Defined Networking (SDN) stands out as a leading-edge technology that empowers organizations to establish a streamlined, secure, and sophisticated network infrastructure. A key advantage of SDN lies in its ability to minimize costs and infrastructure overhead. Among these threats, distributed denial of service (DDOS) attacks are considered as dangerous and common. In this work, DDOS attacks are addressed effectively through machine learning techniques. A random forest method is proposed for DDOS mitigation. The proposed system’s efficiency is evaluated through comprehensive experiments using a dataset comprising diverse network traffic types, including Brute Force Intrusion, Benign traffic, SQL Injection, XSS Intrusion, and DDOS. The experimental work was carried out using MATLAB, with a detailed analysis conducted based on factors which includes Recall, F-measure, and Precision. The results obtained underscore the outstanding accuracy of the proposed random forest algorithm, achieving an impressive accuracy rate of 99.5%.