Projects:2017s1-101 Classifying Network Traffic Flows with Deep-Learning

Introduction

Deep-learning has gained strong prominence in recent years in a range of application areas, including image classification, textual detection, speech recognition, self-driving cars, and was successfully adopted by IBM and Google for their artificial intelligence projects in winning competitions such as Chess and Go. This project involves developing and applying deep learning techniques for classifying network communications traffic.

The Internet consists of a wide variety of network application traffic. Identifying these applications and understanding how their traffic flows behave is essential for monitoring and management of networks and their infrastructure.

Project Team

Clinton Page

Daniel Smit

Kyle Millar

Supervisors

Dr Cheng-Chew Lim

Dr Hong-Gunn Chew

Dr Adriel Cheng (DST Group)

Objectives

- Gain knowledge about the application of deep-learning for classifying network traffic flows

- Conduct experiments on synthetic traffic flows and/or make use of communications flow data from real-life enterprise networks

- Develop network traffic classifying software using deep-learning techniques to an acceptable accuracy when comparing against the results of previous years