Thermal Imaging Decision Applicable to TEMPEST?

[David Koontz]

>      The Supreme Court's decision against thermal imaging appears 
>      to be applicable to TEMPEST emissions from electronic devices.
>      And is it not a first against this most threatening vulnerability
>      in the digital age? And long overdue.
> 
>      Remote acquisition of electronic emissions, say from outside a 
>      home, are not currently prohibited by law as far as I know. And
>      the language of the thermal imaging decision makes it applicable
>      to any technology not commonly in use.
> 
>      Conventional wisdom of security wizards are that the emissions
>      are very difficult to acquire from more than a hundred yards or
>      so, but James Bamford claims in his recent "Body of Secrets" that
>      NSA was able to acquire leaky emissions from Russian crypto 
>      equipment 6 miles offshore Cuba in the 1960s. Advances in 
>      technology would presumbably increase that capability.

How about EMI protection reducing the ability, instead?

I recall shortly after leaving tech school, running accross a
KW-26B/TSEC
(60 Volts peak to peak and 60 ma current loop teletype encryption
system,
capable of operating between 37 point something and 55 baud (bits per
second)).
Having been trained on C and D versions it was interesting to see the
difference.
The C and D versions were MILSTD-188B compliant (12 V PP low current,
compatible
with RS-232C/RS-422), had a line filter module installed in the case by
depot
level maintenance, modified output driver cards and eventually included
a solid 
state relay instead of one with make/break contacts.  I've also applied
some RED 
side low current mods to KW-7/TSEC tactical teletype encryption boxes
(the KW-7 
used an ancillary filter module for non-tactical installations).  

The reason for this was the TEMPEST program.  A 60 V signal high current
loop
is intended to drive a teletype signal along extremely long wire
landlines.
You get inductive effects from the wire line and the relay switching
much as
you do the ignition of internal combustion engine.  I've also read the
Bamford
book recently, and am not at all surprised that it a teletype signal
could
be detected at that kind of distance.  The TEMPEST briefing we received
in
tech school talked about detecting teletype signals at large distances,
similar
to the Black Crow guys in Southeast Asia killing trucks on the Ho Chi
Minh trail 
by detecting and tracking ignition systems.  

The ability to detect these emissions leads to the desirability of
RED/BLACK
separation.  If you could detect the enciphered teletype link at a
considerable
distance, mostly because of the wire length, it is nice if you don't
have
classified information coupled on it.  This leads to other things, long
not
allowing anything that can be an RF generator into a comm center,
unplugable
or isolated phones with separate ringers, using ferrous conduits
(separate
for RED/BLACK), RED and BLACK power separation, and filtering between
the
two environments.  I recall the first time I saw and LED based
optoisolator,
it was used in the control system of a Narrow Band Secure Terminal
(NBST) 
secure voice terminal, signaling the secure switch from the RED side
phone
to relays that switched line connections from non-secure voice to a
MODEM,
and subscriber side connections from the line to a vocoder.

I've worked on KW-26C/TSEC number 10 (it had a christening plate with
the month 
and year I was born), and KG-13/TSEC number 1 (non anodized chassis, in
a classroom
at Lackland).  The Air Force installed TEMPEST mods for all their crypto
in the
early 70s.  I recall a directive to inspect all the crypto mods in the
mid 70s.



---------------------------------------------------------------------
The Cryptography Mailing List
Unsubscribe by sending "unsubscribe cryptography" to majordomo at wasabisystems.com