On Wednesday, March 12, 2003, at 02:40 PM, Thomas Shaddack wrote:
The closest encounter I had with superconductors was when I was
helping a
friend with some measurements on some uranium-based ceramics. Was both
brief and nice, and I lost fear of liquid nitrogen there.
Rational fear of LN is a good thing, though. Minor splashes aren't bad,
but enough can cause serious burns.
I also worked with uranium in ceramics, though they were not
uranium-based (though sometimes we thought they were!).
Jamming is grossly less efficient than detection. If you want an
explanation, let me know and I'll spend 10 minutes writing a small
piece on it. But first, think deeply about why this is so. Think
especially about recovering signals from noise.
Had my brush, though only theoretical, with integrating repeating
signals
back at school, when I was learning how to interpret NMR spectrums and
how
they work. (Good old times, it feels like yesterday.)
Sorry, hadn't specified I am not talking about RFID tags anymore; was
thinking about at least partially alleviating/sidestepping the problems
with shielding of standard desktop computers.
But will be definitely interested in the minilecture.
Sounds like you already have the gist. There are many good ways to pull
weak signals out of noise, either by direct integration over time or by
chopper techniques (e.g., only looking in narrow time intervals, via
gated integrators and boxcar averagers).
And if the RF ID tag is sending out a signal over a couple of different
frequencies, using some pseudorandom sequence for the
frequency-hopping, then the noise gain can be enormous. That is, an
attacker trying to jam a spread-spectrum (Direct Sequence Spread
Spectrum, DSSS, typically) signal will have to match and greatly exceed
the frequences and times.
Even better, pulse systems which send out ultrawideband signals at
various coded time points (so-called Gold codes, or Kosami codes, for
example) are even more difficult to jam.
You mention that your point was about jamming intercepts from
insufficiently-shielded computers, a la TEMPEST, which, by the way, is
not an acronym (To Ensure More Private Eavesdropping-Safe
Telecommunications--NOT).
This is similar to trying to hide phone bugs by running a background
noise source, like a shower or a television set. But as with those
attempts, a skilled eavesdropper can strip out nonrandom noise
sources like music or television, thus improving S/N ratios. Or the
quasi-random noise of a shower just adds to the baseline of noise
already present. (And multiple detectors can help in various ways, much
the way noise-cancellation headphones work...off the shelf consumer
technology, so imagine what the spooks have.)
More dB of eavesdropping attenuation is gotten by reducing the signal
than by increasing the noise, short of the equivalent of jet engines.
Better to whisper than to speak normally but turn on cover noise
sources elsewhere in a room.
Measuring leakage at a distance of a few centimeters is easy to do. And
if a leakage signal is very, very small at a few centimeters, the usual
inverse-square falloff will make it truly tiny at 100 meters or so.
(Where a van might be parked outside one's flat, for example.)
I'm not saying RF emissions are not an issue. Much was written about
this some years ago, even here on this list, when Van Eyck Radiation
(just the RF) was being studied. Ross Anderson at Cambridge and his
group have been doing lots of work on this.
Truly sensitive communications may be best done on laptops, even
laptops in metal mesh bags. (Either with one's head poked into the bag,
or a bag big enough to enclose the user and laptop, etc.)
There are also heads-up LCD displays now costing less than $600, which
can be used with handheld computers and the like. Besides (likely, but
don't quote me) low emissions from the start, a mesh hood would be very
easy to construct, thus knocking probably another 30 dB off the already
low emissions.
Note that the inverse-square law falloff and the vast number of
communications is probably why the Osama Bin Laden deputy, Sheik
Mohammed, wasn't caught because of RF emissions from his laptop, but
instead because of an informant (as I understand things).
I would strongly bet on quietness of computers winning out over
increased RF detection capabilities. (Needless to say, detection goes
as the square of the antenna size, so even really large antennas don't
have that many dB of extra capture capability, compared to quietness at
the source.)
--Tim May
The only purpose for which power can be rightfully exercised over any
member of a civilized community, against his will, is to prevent harm
to others. His own good, either physical or moral, is not a sufficient
warrant. --John Stuart Mill
