[tor-dev] Simulating how Laplace noise added to directory-request statistics would affect user number estimates

[Karsten Loesing]

Hi everyone,

we'd like to improve directory-request statistics by obfuscating values
on relays before they are reported to the directory authorities.

A possible obfuscation method is to add Laplace noise to request counts
for all ~250 countries, so that it's unclear whether a request was
actually made by a user or is just noise.  In fact, we did a similar
thing to onion service statistics two years ago.

But before we do this we need to find out whether obfuscated values
would still be useful enough to estimate user numbers in the Tor network.

We ran a simulation using archived descriptors and put our results
including a graph, CSV files, and simulation code on this wiki page:

https://trac.torproject.org/projects/tor/wiki/org/teams/MetricsTeam/ObfuscationSimulationAnalysis

The result is that we can't just go ahead and put in this noise, because
we'd have to improve our user number estimation algorithms first.
Otherwise we might risk losing one of our most important statistics: the
number of daily Tor users by country.

If you have any thoughts on these results or want to help make the
simulation more accurate, please let us know!

All the best,
Karsten (for the metrics team)

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 495 bytes
Desc: OpenPGP digital signature
URL: <http://lists.torproject.org/pipermail/tor-dev/attachments/20170519/67dd24a3/attachment.sig>