Randy's Work Blog

Thomas Holterbach, Emile Aben, Cristel Pelsser, Randy Bush, & Laurent Vanbever; Measurement Vantage Point Selection Using A Similarity Metric; Applied Networking Research Workshop (ANRW 2017)

In a measurement platform with a wide selection of vantage points, it can be challenging to select the most appropriate points to source measurements from. One example of such platform is RIPE Atlas [2] which currently hosts over 9600 active vantage points, which can be selected based on categories, such as origin AS or country. When setting up a measurement, users are limited in how many vantage points they can use. This is not only due to limitations that the mea- surement platform imposes, but collecting data from a large number of vantage points would mean a large volume to analyse and store. It therefore makes sense to optimize for a minimal set of vantage points with a maximum chance of observing the phenomenon in which the user is interested.

Network operators may need to debug a network service with only limited information about the problem (“Our network is slow for users in France!”). A diversity metric would allow selection of the most dissimilar vantage points, in an attempt to explore the network phenomenon from as diverse angles as possible. If one nds an interesting network phenomenon, one could use the similarity metric to advantage by selecting the most similar vantage points to the one exhibiting the phenomenon, in an attempt to validate the phenomenon from multiple vantage points.

We propose a novel means of selecting vantage points, which is not based on categorical properties (such as origin AS, or geo- graphic location), but rather on the topological (dis)similarity be- tween vantage points. We describe a similarity metric across RIPE Atlas probes, and show how this performs better for the purpose of topology discovery than the default probe selection mechanism built into RIPE Atlas.