Measuring the evolution of internet peering agreements

  • Authors:
  • Amogh Dhamdhere;Himalatha Cherukuru;Constantine Dovrolis;Kc Claffy

  • Affiliations:
  • CAIDA;Georgia Tech;Georgia Tech;CAIDA

  • Venue:
  • IFIP'12 Proceedings of the 11th international IFIP TC 6 conference on Networking - Volume Part II
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

There is much interest in studying the structure and evolution of the Internet at the Autonomous System (AS) level. However, limitations of public data sources in detecting settlement-free peering links meant that prior work focused almost exclusively on transit links. In this work, we explore the possibility of studying the full connectivity of a small set of ASes, which we call usable monitors . Usable monitors, while a subset of the ASes that provide BGP feeds to Routeviews/RIPE collectors, are better suited to an evolutionary study than other ASes.We propose CMON , an algorithm to classify the links of usable monitors as transit or non-transit. We classify usable monitors as transit providers (large and small), content producers, content consumers and education/research networks. We highlight key differences in the evolution of connectivity of usable monitors, and measure transitions between different relationships for the same pair of ASes.