HybridDAD: Detecting DDoS Flooding Attack using Machine Learning with Programmable Switches

Mostafa Roshani, Mehdi Nobakht

2022Proceedings of the 17th International Conference on Availability, Reliability and Security15 citationsDOI

Abstract

Recently, Distributed Denial of Service (DDoS) attack has become more disruptive and detrimental in businesses, healthcare and government sectors. DDoS prevents legitimate users from receiving expected services from service providers, e.g, ISPs, and web servers. Data plane programming using P4-enabled devices has recently come to attention to detect and mitigate DDoS attacks as an in-network defense. Moreover, integrating this technology with machine learning techniques has been emerged in recent years. A few research work propose to use P4 and ML models to detect DDoS attacks. However, to best of our knowledge, they aim only on a single attack type despite the fact that, according to a new research study, the attackers often conduct various attack types in a short period of time.

Topics & Concepts

Denial-of-service attackComputer scienceApplication layer DDoS attackTrinooComputer securityServerFlooding (psychology)HoneypotComputer networkThe InternetWorld Wide WebPsychotherapistPsychologyNetwork Security and Intrusion DetectionAdvanced Malware Detection TechniquesInternet Traffic Analysis and Secure E-voting