ML Based Detection & Classification of High- and Low-Rate DDoS Attacks

ML  Based Detection & Classification of High- and Low-Rate DDoS Attacks

M.Jagadeesh1, R.Dhileep Kumar2, K. Kowsthubha3, V.Shanmuka Rao4, Sk.Althaf Rahaman5

[1-4]B.TechStudent, [5]AssistantProfessor,LIET[1,2,3,4,5]ComputerScienceandSystemEngineering,LendiInstitute of Engineering and Technology, Viziangaram.

---------------------------------------------------------------------***---------------------------------------------------------------------

ABSTRACT

The Machine Learning-Based Detection and Classification of High-Rate (HR) and Low-Rate (LR) DDoS Attacks system is designed to strengthen network security by identifying and categorizing different types of attack traffic. This intelligent detection framework leverages machine learning techniques to analyze network behaviour and distinguish between legitimate and malicious requests. The system utilizes the XGBoost classifier to accurately classify HR attacks, which generate overwhelming traffic surges causing immediate disruptions, and LR attacks, which operate subtly to evade traditional detection methods. A comprehensive dataset containing both attack types is used to train the model, ensuring high detection accuracy. The system features a scalable and efficient architecture capable of real-time traffic monitoring and classification. With secure data handling, automated analysis, and adaptive threat detection, this approach enhances cybersecurity defenses, minimizes downtime, and provides a proactive solution to mitigating evolving DDoS threats.

Key Words: DDoS Detection, High-Rate (HR) Attacks, Low-Rate (LR) Attacks, Machine Learning, XGBoost Classifier, Network Security, Traffic Analysis, Attack Classification, Real-Time Threat Mitigation, Cybersecurity Defense.