Machine-Learning based DDoS attack detection in SDN

2018 COE Engineering Design Project (BM03)

---

Faculty Lab Coordinator

Bobby Ma

Topic Category

Networking

Preamble

Software Defined Networking (SDN) is an emerging network technology. In SDN, a centralized controller is used to control data flows and manage network policy. Distributed Denial of Service (DDoS) attack is the most popular type of network attacks for disabling network services. In this project, students will learn the structure of SDN and program the controller for the detection of DDoS attack. In addition, students will evaluate several machine learning algorithms based on their detection performances.

Objective

To implement a DDoS attack detection system in the SDN platform. To evaluate various Machine Learning algorithms based on their detection performances.

Partial Specifications

1. Study the concepts of SDN, DDoS and machine learning.
2. Use the OpenDayLight Controller and Mininet to setup a SDN network.
3. Use Hping3 or other open-source attacking tools to generate attack traffic.

Suggested Approach

The project has three stages. In the first stage, students will setup a SDN network using OpenDayLight Controller and Mininet. In the second stage, students will learn how to program the controller to control the traffic flow in the SDN network. In the third stage, a machine-learning based DDoS detection system will be implemented and tested.

Group Responsibilities

It is expected that all the members of the group are involved in the research, development and implementation of the project.

Student A Responsibilities

1. Study and understand the concepts of SDN, DDoS attack and Machine Learning. 2. Implement a Machine learning-based DDoS attack detection system in SDN by programming the SDN controller. 3. Evaluate various machine learning algorithms based on the detection performances

Student B Responsibilities

1. Study and understand the concepts of SDN, DDoS attack and Machine Learning. 2. Implement a Machine learning-based DDoS attack detection system in SDN by programming the SDN controller. 3. Evaluate various machine learning algorithms based on the detection performances

Student C Responsibilities

1. Study and understand the concepts of SDN, DDoS attack and Machine Learning. 2. Implement a Machine learning-based DDoS attack detection system in SDN by programming the SDN controller. 3. Evaluate various machine learning algorithms based on the detection performances

Course Co-requisites

COE 768