Detection of Cloud Network Traffic Abnormalities

This research project aims to develop a technique for detecting and analyzing security incidents in their early stages, reducing the potential impact on an organization’s operations. Conventional methods of deep packet inspection (DPI) and network monitoring solutions only identify frequently occurring traffic patterns, and security threats are often not detected until it’s too late. The project investigates a new approach that takes a “horizontal perspective” to detect outliers by identifying packets with out-of-distribution attributes and a “vertical perspective” to detect unusual patterns formed by common packets during a certain interval. The project will also develop countermeasures for specific attacks and general abnormalities. The expected benefit for the partner organization is early detection of security incidents, reducing the risk of data breaches and damage to the organization’s operations.

Faculty Supervisor:

Dehan Kong

Student:

Partner:

SOTI Inc

Discipline:

Computer science

Sector:

Information and cultural industries; Professional, scientific and technical services

University:

University of Toronto

Program: