[tor-talk] Exit Traffic classification and discrimination

[Fabio Pietrosanti (naif) - lists]

On 2/2/16 1:50 PM, Roger Dingledine wrote:
> My suggestion to him at the time was to write up the details for why his
> design is safe, and get them vetted by other researchers in public, before
> proceeding. I haven't talked to him about how it's going since then.
> Hopefully he didn't skip too many of the recommended steps. :)

Another very important challenge would be, how to discriminate those
traffic, due to the CDNs and other geo-dependent complexity.

How to get Alexa top-500 (that we may expect to unlikely be
abuse-generating sites), get their IP addresses/netblocks and route them
over a specific Tor Exit IPs address to measure if:
- They generate abuses
- How much of the Tor Exit traffic they generate

That would be a simple, yet privacy safe, way to evaluate if those
traffic classification and discrimination approach could work in
*uber-simplifying* the life of a Tor Relay operator.

-- 
Fabio Pietrosanti (naif)
HERMES - Center for Transparency and Digital Human Rights
http://logioshermes.org - https://globaleaks.org - https://tor2web.org -
https://ahmia.fi