[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: TEMPEST (fwd)




Forwarded message:

> Subject: Re: TEMPEST
> Date: Tue, 10 Feb 1998 18:32:28 +0000
> From: Markus Kuhn <[email protected]>

> You might want to consider to read van Ecks paper yourself:
> 
>   Wim van Eck: Electromagnetic Radiation from Video Display Units: An
>   Eavesdropping Risk? Computers & Security 4 (1985) 269--286
> 
> If you don't have C&S in your library, you might also find a
> scan on <http://www.eskimo.com/~joelm/tempest.html>

I read the paper when it came out, but that's been a while now..:)

Thanks for the reference, I'll take a look at it over the next few days.

> > As I alluded to with the comment I made about integrating the received
> > signals. The easiest way to do this sort of stuff with a small budget is
> > with a flying capacitor integrator.
> 
> Please explain! I was more thinking in terms of digitizing everything
> and solving large per-pixel systems of equations to separate the drifting
> images, which does not sound much like something equivalent to a
> single hardware integrator, but more like something that keeps a
> workstation very busy for a few minutes.

I'll describe how it works in hardware and then how I've done it in
software.

Hardware:

You have a bank of n capacitors that are hooked to a amplifier via a
mux. You use a counter that you start with some sort of trigger that
sequentialy connects each cap in turn to the amplifier. The charge and as
a consequence the voltage varies depending upon whether the new voltage is
higher or lower than the last sample. The assumption being that the signal
itself is time invariant and that the variations in signal strength are
noise and random in nature. After only 3-4 cycles through the cap bank you
begin to pull a workable signal out of the noise. The nice thing is that
this works for signals that are below the noise floor and buried in the
grass.

In a VanEck monitoring situation you'd need a bank of n capacitors where n
is the number of horizontal pixels. You will also need to array each bank
into a m row matrix where m is the number of scan lines. You can get around
some of the bulk by using a gyrator circuit to synthesize the capacitor and
some sample-and-holds to synthesize the array.

The advantage here over software is that this signal can be sampled as its
built in real-time (use instrumentation amps) unless you happen to have
dual-port ram in your computer (I've never had access to such hardware) and
fed to whatever equipment you want to process it.

Software:

You take a A/D converter and drive an array with it. You sync it the same
sort of way as the hardware methods above. Each element in the array
contains the current estimate for that particular sample of the total
waveform. Each time you come back to the same array element you compare the
last value to the new value. You take the value that is half-way between the
two and store that as the new array value for the next cycle. In short order
a signal comes out.

Either the hardware or software method will naturaly integrate the signal
within the sample window because of the capacitor effect. This compensates
for clock and signal jitter quite nicely.

This method will require a D/A conversion process if you want to feed it
to any external equipment. Though it should be perfect for Van Eck
monitoring. Simply use your video frame buffer as your sample storage array.

Enjoy.


    ____________________________________________________________________
   |                                                                    |
   |                   The obvious is sometimes false;                  |
   |                   The unexpected sometimes true.                   |
   |                                                                    |
   |                                       Anonymous                    |
   |                                                                    |
   |                                                                    | 
   |            _____                             The Armadillo Group   |
   |         ,::////;::-.                           Austin, Tx. USA     |
   |        /:'///// ``::>/|/                     http://www.ssz.com/   |
   |      .',  ||||    `/( e\                                           |
   |  -====~~mm-'`-```-mm --'-                         Jim Choate       |
   |                                                 [email protected]     |
   |                                                  512-451-7087      |
   |____________________________________________________________________|