On Power-Law Relationships of the Internet Topology

This paper was a refreshing diversion from many of the protocol and systems building papers we have been reading. They grabbed three graphs, each separated by 6 months, of the Internet from the National Laboratory for Applied Network Research and plotted some simple statistics using log-log scale and noticed a lot of straight lines. Ha.

Axis of each powerlaw:

1. <node rank (based on # of fanout connections from node), fanout>


2. <node rank, frequency (bucketing fanout)>


3. <# of nodes within range, radius of range (# of hops)>

Are #'s 1 and 2 very different?