From: tcmay@got.net (Timothy C. May)
Date: Sun, 5 Nov 1995 01:02:29 +0800
To: Tim Philp <bplib@wat.hookup.net>
Subject: Re: video as a source of public randomness
Message-ID: <acbf861a06021004706b@[205.199.118.202]>
MIME-Version: 1.0
Content-Type: text/plain


At 2:17 PM 11/3/95, Tim Philp wrote:
>        Would not generating random numbers using video sources be
>suseptable to the same 'external' influences as generating random numbers
>from radio static? External RF fields could skew the 'randomness' of the
>data.

Do a thought experiment, if not an actual experiment.

Tune your t.v. to an unused channel. Use no antenna, or perhaps just small
rabbit ears. Notice the swirl of snow (as in "Snow Crash"). Imagine
sampling bits from this snowstorm, perhaps 10 per frame. Apply the usual
tricks to remove biases in one direction or another (*) and apply hashes as
desired to distill bits of entropy.

(* Such as von Neumann's scheme to turn even biased coins into "fair"
sources: toss a coin twice and let H T be a 1 and T H be a 0.)

To complete the thought experiment: Imagine the difficulty of an attacker
being able to module this snowstorm in any predictable way. Imagine the
difficulty when the antenna input has been disconnected, or the rabbit ears
moved by air currents in the room, or....

"All crypto is economics." The cost of an attack based on modulating this
source of randomness would be in the tens of billions of dollars, or more,
and likely impossible at any cost at this time. (Depending on the sampling
details, the antenna input, etc. I'm not saying a black bag job could not
be done to alter the inputs to the t.v., just that external RF manipulation
is unlikely in the extreme to be economically feasible.)


>        As another thought, has anyone done any work on RNGs involving
>chaotic processes such as fluid dynamics and turbulent flow? I suspect
>that pressure, or other parameter variences, in turbulent flow could yield
>good random numbers. As a bonus, these parameters are easily measurable
>without special, exotic, equipment and should be inexpensive

Ignoring the work allegedly done on "chaotic encryption," the RF method
described above essentially is relying on turbulence: the atmospheric
variations ("spherics," "whistlers," are some of the buzz words) affect the
snowstorm. Also, the rabbit ear antennas move in response to room air
currents.

Again, all unpredictable. (And as I noted above, sampling strategy is
important...)

I  don't plan to belabor this point. Radioactive decay sources are
certainly fine, though not likely to be purchased by most people. (I have
nothing against radioactive decay, as some of you may know--it made my
career.)

--Tim May


Views here are not the views of my Internet Service Provider or Government.
---------:---------:---------:---------:---------:---------:---------:----
Timothy C. May              | Crypto Anarchy: encryption, digital money,
tcmay@got.net  408-728-0152 | anonymous networks, digital pseudonyms, zero
Corralitos, CA              | knowledge, reputations, information markets,
Higher Power: 2^756839      | black markets, collapse of governments.
"National borders are just speed bumps on the information superhighway."