[time-nuts] GPS jamming susceptibility

[jimlux]

John Green wrote:
> 
> jamming anyone's GPS. A while back, I was looking at one of those
>  It doesn't look capable of putting out
> more than 50 milliwatts or so into a 2 inch antenna

  The GPS antenna is perhaps 35 feet away
> with a cinder block wall, a brick wall, and a metal roof in between. I
> also put a 15 Db attenuator between it and the antenna with almost the
> same result.

> down. Has anyone here had any actual experience testing GPS receivers
> for susceptibility?
> 


OK... typical received signal at a GPS receiver (L1) is on the order of 
-130dBm.  Thermal noise floor (assuming noiseless receiver and no 
losses) is -114 dBm in 1 MHz BW.

Remember, the typical GPS is a single bit quantizer, which works just 
fine considering the signal is 20dB below the noise floor.

So, let's do a little link calculation: 32+20log10(1500)+20log10(.010) 
between isotropic antennas (which is not a bad starting point for your 
jammer and the GPS)..
32+64-40 -> a link loss of 56 dB.. you're radiating +17dBm, so let's 
call it -40dBm into the GPS..  Yep, jamming is almost assured..

But at that kind of power, you'd jam almost ANY receiver that's trying 
to receive a signal at -130dBm.  90dB instantaneous dynamic range is 
pretty good when you can't use a filter to remove the interfering signal 
(e.g. a HF receiver has a narrow band filter in the IF to solve this 
problem).

realistically, you need about, say, 10 dB J/S so you'd need -120 dBm 
into the receiver from the jammer.  A microwatt 10 meters away would do 
it nicely.  Inverse square helps a bit.. if you were 1 km away, your 
interfering 50mW signal would be down another 40 dB.. -80dBm.

10km away, your jammer is down into the area where it probably won't jam 
all the time.


Obviously, *real* radios that need reliable GPS reception do things to 
make life easier.  Aside from using 1.5-2 bit detection, or signal 
excisers, etc.  There are also techniques that rely on looking at the 
post correlation signal (where the interferer is suppressed to a certain 
extent): with modern signal processing, you can correlate against all 
possible phases of the code in one shot, for instance.