DDoS Detection using Machine Learning Techniques

Abstract

A Distributed Denial of Service (DDoS) attack is a type of cyber-attack that attempts to interrupt regular traffic on a targeted server by overloading the target. The system under DDoS attack remains occupied with the requests from the bots rather than providing service to legitimate users. These kinds of attacks are complicated to detect and increase day by day. In this paper, machine learning algorithm is employed to classify normal and DDoS attack traffic. DDoS attacks are detected using four machine learning classification techniques. The machine learning algorithms are tested and trained using the CICDDoS2019 dataset, gathered by the Canadian Institute of Cyber Security. When compared against KNN, Decision Tree, and Random Forest, the Artificial Neural Network (ANN) generates the best results.