On Mon, 22 Dec 2003 23:08:58 -0800, pervect
wrote:


I'm really not sure how quickly you can count on taking out a spread
spectrum transmitter. Especially when it's put on a low duty cycle
transmit mode rather than a continuous transmit mode.

I'm going to throw some numbers at this problem.

The shannon-hartley capacity of the communication channel should be

B(log2(1+S/N)), where S/N is the signal/noise ratio (measured at the
receiver), and B is the bandwidth.

Let's say our goal is to have the same channel capacity as a 25khz
channel with a 10 db S/N. That would be about 86khz. Round it up to
100khz, this is just a BOTE calculation.

Now lets suppose our spread spectrum channel is about 10Ghz wide.

log2(1+S/N)= 10^-5

S/N=.69e-5 (needed at the receiver)

Assuming inverse square law propagation, we'll have to be about
1/sqrt(.69e-5) = 400x closer to the source than the receiver is to get
a S/N of 1. So if the reciever was 40km away from the transmitter,
we'd have to be within about 100m of the source to have a S/N of 1.

With long enough integration times from a fixed site, we can probably
get some sort of bearing with a S/N 1, but I doubt that any sort of
rapidly moving radiation seeking missile is going to be able to lock
on unless the signal is at least as strong as the noise. It should
also be pretty easy to setup false antennas transmitting low levels of
broadband noise to make any such missile's job very difficult unless
the attacker doesn't mind launching a bunch of them and also doesn't
care what they might hit (collateral damage).