CORIA 2.0 – Augmenting a Universal Framework for Connectivity Risk Analysis

Abstract

Using graph structures one can model various types of problems. The Internet is one of them. CORIA is a multi-purpose framework to analyse connectivity risks of graph structured data which was designed using Internet topology data. We augment the existing framework for graph analysis CORIA with the objective to increase scalability and broaden the feasible range of application.

To this end we transform CORIA such that it functions as a cloud-based tool that can store and display results for several graphs simultaneously thus making it suitable for the concurrent use by multiple researchers, we adopt faster algorithms, reducing metric calculation times by at least 58%, and we implement modules for the classical dynamic analysis of fragmentation under various deletion modes and for graph visualization. In doing so we review the Unified Risk Score (URS) and confirm the general underlying assumptions while suggesting alterations to the relative weight of components.

Owing to recent political developments we also simulate the disruption caused to the Internet topology on the Autonomous System level by cutting off all Russian systems as an example use case, recording a deterioration of the global AS level graph.

Keywords: graph analysis, robustness, fragmentation, Internet, autonomous systems, AS, vulnerability, network