Automating morphing DDoS attacks with ML techniques

In recent years, Distributed Denial of Service (DDoS) attacks have evolved, becoming more sophisticated and challenging to mitigate. A novel form, morphing DDoS attacks, dynamically changes attack vectors to evade defenses. This project aims to explore adaptive morphing DDoS attacks, which adjust sent packets based on real-time data about the target network.

Objectives

The aim of this project is to improve our current configuration for the execution and analysis of Adaptive Morphing DDoS Attacks. On top of the existing project, the student should:

  • Extract new network metadata features for the attack
  • Create an ML classifier to detect whether or not a DDoS defence is being used.
  • Carry out the experiments and compare the results with the existing project.

Requirements

  • You should be familiar with Java, Python, and Shell scripts.
  • Ideally, you have some knowledge of Network metrics and DDoS attacks.

This research project is aimed at one MSc student. The student will work with Eric Jollès (EPFL) and Dr. Muoi Tran (ETHZ).

Supervisors