Mobility

On the Success of Network Topology Inference Using a Markov Random Walk Model for Nested Routing Policies

Free registration required

Executive Summary

Network topology inference is a topic of ongoing interest, as researchers continue to try to identify the topology of the internet, other networks, and the people own networks for maintenance purposes. In this paper, the authors explain why a simple algorithm based on network co-occurrence measurements and a Markov random walk model for routing enables perfect topology reconstruction, despite the seeming model mismatch to real network routing. They show that topology measurement and inference based on this model allows detection of routing bugs, mis-configuration, or even routers that deliberately misreport routing data.

  • Format: PDF
  • Size: 660.77 KB