Index of /datasets/2010-as-annotations-data-supplement

Icon  Name                            Last modified      Size  Description
[DIR] Parent Directory - [   ] README.pdf 05-Mar-2010 17:24 280K [TXT] README.txt 01-Jun-2017 15:49 2.5K [TXT] as-rel.a0.01000.20091006.txt.gz 05-Mar-2010 17:24 608K [TXT] copyright.txt 05-Mar-2010 17:24 1.6K [TXT] links.20091006.txt.gz 05-Mar-2010 17:24 306M [TXT] nodes.20091006.txt.gz 05-Mar-2010 17:24 170M [TXT] nodes.as.20091006.txt.gz 05-Mar-2010 17:24 66M [TXT] prefix2as.20091006.txt.gz 05-Mar-2010 17:24 1.2M
The following datasets constitute the dual-graph representation of the Internet
using data collected in October 2009:

nodes.20091006.gz ( node dataset)
is the set of routers inferred by kapar. For each router, the node dataset
contains the set of interfaces that were inferred as being on that router.

File format:
  node <node_id>:   <i1 >   <i2 >   ...   <in>
indicates that kapar inferred the node with id node_id to have the n interfaces 
i1 to in

links.20091006.gz ( link dataset)
contains, for each link, the set of routers and router interfaces that kapar inferred
as sharing that link.

File format:
  link <link_id>:   <N1>:i1 <N2 >:i2 [<N3 >:[i3]]   ...   [<Nm>:[im]]
indicates that kapar found a link link_id connecting nodes N1 to Nm . Note that
kapar always determines at least two nodes and the associated interfaces for
each link. There may, however, be additional nodes on the same link due to
the presence of LANs or clouds with multiple attached interfaces.

By joining the node and link datasets, one can obtain for each router a set of
known interfaces and inferred interfaces. Known interfaces were measured directly
and are in the node dataset; Inferred interfaces result from entries in the
link dataset that indicate that router r1 has a link to interface i2 on router r2,
but we did not see an actual interface on router r1 . The interfaces on an IP
link are typically assigned IP addresses from the same prefix, so we assume
that router r1 must have an inferred interface from the same prefix as i2.

prefix2as.20091006.gz ( IP-to-AS)
contains the list of IP prefix to origin AS found in the BGP dumps.

File format:
  <prefix>   <length>   <origin AS>
When multiple ASes were found as the origin of the same prefix, they are
concatenated with an underscore.

nodes.as.20091006.gz ( AS assignment dataset)
contains, for each router found in the node dataset, the AS that is
the inferred owner of that router according to to our final
Election+Degree assignment heuristic.

File format:
  <node_id>   <AS>   <method>

as-assignment.tgz ( AS assignment code)
contains the code to generate the AS assignment dataset.

as-rel.a0.01000.20091020.gz ( AS relationship dataset)
contains the business relationship associated with each link in the BGP data,
computed (see above) using the AS relationships algorithm by
Dimitropolous, et al.

File format:
  <AS1 >   <AS2 >   <rel>

where
  rel is -1 if AS1 is a customer of AS2
  rel is  1 if AS1 is a provider of AS2 
  rel is  0 if AS1 is a peer of AS2 
  rel is  2 if AS1 is a sibling of AS2