Realtime detection and visualization of BGP anomalies using machine learning
Victoria University of Wellington
The number of Border Gateway Protocol (BGP) anomalies, such as hijacking, is growing due to limited detection capabilities. BGP hijacking maliciously reroutes Internet traffic, causing Denial of Service (DoS) to major Internet Service Providers (ISP) or redirection attacks to Internet users. Ensuring that the Internet functions with minimal disruptions is critical.
This proposal aims to use innovative approaches in network modelling and machine learning to detect and pre-empt any impending anomalous events. The project is led by Victoria University of Wellington (VUW, New Zealand) collaborating with Inner Mongolia University (IMU, China) and Prince of Songkla University (PSU, Thailand). The project aims to build on the foundational research by VUW to develop a realtime system that enables network administrators to detect and visualize impending anomalous events before they occur, thus allowing them to take prompt remedial actions.