Inter-Domain Access Volume Model: Ranking Autonomous Systems

Exploring topological structure at Autonomous System (AS) level is indispensable for understanding most issues in Internet services. Previous models of AS graph involve address or connectivity information separately, and neither of these information can serve as a comprehensive metric for evaluating an AS’s contribution to the global routing. In this paper, we propose a new model for AS ranking named IDAV (Inter-Domain Access Volume). IDAV introduces the quantity of routed addresses into AS graph, and enriches the methodology of Internet marco structure inference. In IDAV, the magnitude of AS is measured with the primary eigenvector of the access volume matrix (Carriage Matrix) for AS graph. We construct the AS graph by parsing the forwarding information bases in border gateways. The computation, compared with previous approaches, demonstrates that IDAV model results in more accurate AS rankings. We believe the IDAV model is promisingly useful for studying inter-domain routing and Internet service behavior.