multicast/broadcast experiment at IETF94

IETF Chair <chair@ietf.org> Mon, 02 November 2015 07:03 UTC

Subject: multicast/broadcast experiment at IETF94
Mime-Version: 1.0 (Mac OS X Mail 7.3 \(1878.6\))
Content-Type: text/plain; charset="windows-1252"
From: IETF Chair <chair@ietf.org>
In-Reply-To: <51930CB1-6055-4607-B5B3-52E14A28F5E5@ietf.org>
Date: Mon, 02 Nov 2015 16:03:56 +0900
Content-Transfer-Encoding: quoted-printable
Message-Id: <5A22DA09-28F5-4BE1-9786-07B4EFF3E769@ietf.org>
References: <51930CB1-6055-4607-B5B3-52E14A28F5E5@ietf.org>
To: IETF Announcement List <ietf-announce@ietf.org>
Archived-At: <http://mailarchive.ietf.org/arch/msg/ietf-announce/sGgO_YghjkcK899MWOVb7iah9go>
Cc: Rolf Winter <rolf.winter@hs-augsburg.de>
Precedence: list
Reply-To: ietf@ietf.org, Rolf Winter <rolf.winter@hs-augsburg.de>, IETF Chair <chair@ietf.org>

In Prague we had one experiment that we wanted to run in the IETF
network but were unable to do so due to privacy concerns. The folks
behind the experiment have now redesigned their experiment to
reduce those concerns, and we plan to run it in the IETF network
from Wednesday to Thursday. The affected SSIDs are “ietf” and
“ietf-legacy”.

The experiment and background related to it described further in

http://net.hs-augsburg.de/projects/2015/10/21/ietf-broadcast-revisited.html
https://tools.ietf.org/html/draft-winfaa-broadcast-consider-01

The goal of the experiment is to see whether there is
personally identifiable information in various broadcast
and multicast packets in the IETF network. The general
arrangement is to collect broadcast and multicast traffic,
hash it in a way that allows (probabilistic) comparisons
to be made within the data. The original information 
does not leave the IETF network NOC, and a salted 
hashing process is setup in a manner that does not
allow tracking back to individuals. The process will 
end up losing some information that humans would 
be able to recognise as personally identifying 
information, so the results of the experiment
will only be able to provide a lower bound on the
amount of information.

Rolf will have more detailed description of the process if
anyone is interested.

Jari Arkko

multicast/broadcast experiment at IETF94 IETF Chair