Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ukma!rutgers!att!cbnews!military
From: military@cbnews.ATT.COM (William B. Thacker)
Newsgroups: sci.military
Subject: Re: Spoofing GPS signals?
Message-ID: <6101@cbnews.ATT.COM>
Date: 1 May 89 03:01:43 GMT
Organization: Anasazi Inc, Phoenix AZ
Lines: 49
Approved: military@att.att.com



From: sun!sunburn!mcdphx!anasaz!john
In article <5936@cbnews.ATT.COM> you write:
>
>    I bet the GPS satellite time signals contain error detection
>    codes, if not error correction, which ought to reduce false
>    time output to a minimum, but won't stop a bad person from
>    faking the time.
>
>Does this mean that giving our nuclear missiles GPS-based
>guidance systems makes them vulnerable?  Self-destruct mechanisms
>were avoided because of the possibility of Soviets spoofing the
>signals - what could they accomplish if they could spoof the GPS
>signal, and how feasible is it to do so?

I believe that military GPS signals use spread spectrum transmission.
This means that, except for the simplest spread spectrum scheme, concealing
the spreading sequence is sufficient to conceal the information, and
also to prevent effective spoofing. In other words, the spread spectrum
modulation itself can be used as a form of encryption. Of course,...
if they key got out, that would be bad news. 

    Even if this trick isn't used, modern crypto techniques make it
possible to have a very secure ID and checksum on a message. Even if
the opposition knows how to test for a valid ID, they may not be
able to generate one. For example: use a truly random (quantum effect
determined) sequence generator to generate a random number for each message;
put the resulting number at the start or end of a publicly known
"signature" - It could even be the text name of the satellite with
a high reliability checksum following; Now add the current time
following the random number; Encrypt the whole thing using a trap-door
type block cipher system where only the decryption key is "publicly" known.
(of course, you wouldn't really publish the key - this just means that
 if the decryption key security is breached, the messages still cannot
 be spoofed - only read).

    With a scheme like this, the position can be derived by the normal
method after decrypting and authenticating the message, and the message
cannot be spoofed.

    I suspect a stronger way to attack weapons systems using GPSS is
to jam the satellite signal, or just destroy the satellite. Of course,
the spread spectrum signal is harder to jam if the spread sequence
isn't known to the jammer.

and the resulting numbers
are used as "salt" in a message that also contains the publicly
known