Abstract :
[en] The Internet topology discovery has been an extensive research subject those last years. While the raw data is collected using large traceroute campaigns, additional probing and/or extensive computation are required to gather subsets of IP addresses into single identifiers cor- responding to routers. This process, known as alias reso- lution, leads to a router level map of the Internet.
In this paper, we push further the Internet router level mapping with a new probing tool called MERLIN. MER- L I N is based on mrinfo, a multicast management tool. mrinfo is able to silently collect all IPv4 multicast en- abled interfaces of a router and all its multicast links to- wards its neighbors: it does not need or rely on any alias resolution mechanism. In addition, M E R L I N comes with the advantage of being much more scalable than standard data gathering techniques. In this paper, we deploy and evaluate the performance of MERLIN. We demonstrate that the use of several vantage points is crucial to circum- vent IGMP filtering in order to collect large amounts of routers. We also investigate the completeness of MERLIN by providing a lower bound on the proportion of in- formation that it may miss. Finally, our dataset and the MERLIN implementation are freely available.
Scopus citations®
without self-citations
8