[140437] in North American Network Operators' Group
Re: 23,000 IP addresses
daemon@ATHENA.MIT.EDU (Michael Holstein)
Wed May 11 10:59:22 2011
Date: Wed, 11 May 2011 10:59:08 -0400
From: Michael Holstein <michael.holstein@csuohio.edu>
To: Ken Chase <ken@sizone.org>
In-Reply-To: <20110511143217.GN23277@sizone.org>
Cc: NANOG list <nanog@nanog.org>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
> ("it's one in a billion to crack it! beyond a
> reasonable doubt! we dont have anyone anywhere in our IT who could possibly
> crack it!")
A billion iterations takes what fraction of a second using a high-end
multi-card gamer rig and CUDA? (or for the cheap/lazy, a S3/Tesla instance).
Even for brute-force, although WPA2 is salted with the SSID, 95% of the
time it's still "Linksys". Rainbow tables for the ~140 most common SSIDs
are already available.
I once used GPS and a wifi analyizer to show a map of how large the
possible "cloud" around a standard WRT54G and 2nd floor installation of
the accused's router really was. To make it dumb enough, I used the
pringle's cantenna (literally) instead of a commercial antenna.
The "CSI effect" works when the defense does it too. Juries love to hear
techie stuff these days, it's just that the defense usually can't afford
it. If a sizable community of technical folks were to pro-bono as expert
witnesses, the "presumption of innocence" would return pretty fast.
Cheers,
Michael Holstein
Cleveland State University
