From: Tim May <tcmay@got.net>
Date: Thu, 5 Nov 1998 14:04:53 +0800
To: cypherpunks@Algebra.COM
Subject: Re: TEMPEST laptops
In-Reply-To: <v03130318b2627b816c00@[209.66.100.110]>
Message-ID: <v03130323b266ea53c79e@[209.66.100.110]>
MIME-Version: 1.0
Content-Type: text/plain



At 8:06 AM -0800 11/4/98, Petro wrote:

>	What if, instead of trying to entirely prevent leakage, one did a
>combination of "redirecting" and "masking" emissions.
>
>	Keep in mind I am asking from a point of total ignorance.
>
>	To break the question down further, a tempest attack is limited by
>2 things, distance from the machine (IIRC, the "level" or "strength" of RF
>emissions drops by the square of the distance correct?) and (possibly) the
>presence of other sources of RF in about the same bands.
>
>	Assuming that the signal level drops by the square of the distance,
>then one is far more likely to get tempested from a van outside than an
>airplane overhead correct?  In that case, simply design one of Mr. May's
>brazed copper boxes so that it is open something similar to:
[diagram of semi-open box elided]

Radio waves scatter...they don't just travel in pure line of sight. And
even if they travelled only in line of sight, the reflections from inside
the box and then into the room and then off surfaces....

Microwave ovens work by having the waves bounce around inside a box. Any
significant hole or crack (up to roughly half the wavelength) would let the
waves out.

An open top box will not work.

>The other question is how hard, given a _specific_ machine would it be to
>create a "RF" jammer? Sort of an active defense versus the passive defense
>of a Tempest sheild. build a device that measures the RF coming off a
>machine, and rebroadcasts the opposite (i.e. the negation) of the signal?
>This should, or could "flatten" the signal making it useless.

Unlikely to prevent someone from figuring out what the real signal is. It's
very difficult, generally, to hide a signal with another signal. Noise
won't work, because noise can be filtered or autocorrelated out. A "spoof"
signal can be corrected for.

And we are talking about 100 dB sorts of suppression. Mere factors of a few
with fake signals and noise are meaningless on this scale.

--Tim May

Y2K: A good chance to reformat America's hard drive and empty the trash.
---------:---------:---------:---------:---------:---------:---------:----
Timothy C. May              | Crypto Anarchy: encryption, digital money,
ComSec 3DES:   831-728-0152 | anonymous networks, digital pseudonyms, zero
W.A.S.T.E.: Corralitos, CA  | knowledge, reputations, information markets,
Licensed Ontologist         | black markets, collapse of governments.