[time-nuts] GPS PRN 32

[Magnus Danielson]

From: mikes at flatsurface.com (Mike S)
Subject: Re: [time-nuts] GPS PRN 32
Date: Sat, 09 Feb 2008 21:52:59 -0500
Message-ID: <20080210025300.A4B381165C3 at hamburg.alientech.net>

> At 07:42 PM 2/9/2008, Bruce Griffiths wrote...
> >The Jupiter documentation indicates that PRN range is 1-32 with 0 used
> >to indicate all satellites.
> 
> You should be good, then, since the developers obviously accommodated a 
> PRN 32.
> 
> The statement quoted from the original article ("GPS receivers 
> initially were built to accommodate up to 31 satellite signals...") is 
> incorrect. The PRN range for space vehicles has always been 1-32. (see 
> Table 3-I in 
> http://www.navcen.uscg.gov/pubs/gps/icd200/ICD200Cw1234.pdf ). The 
> concern is that the code in some GPS receivers was poorly written, and 
> set up data structures to support PRNs 0-31, even though PRN 0 is, and 
> has always been, undefined.
> 
> So problems are not a matter of the age of the receiver, but the 
> quality of its programming.

On the 13 December 2006 one operator saw their complete network fail as a
result, which was correlated to the enlargement of the GPS constellation.
The GPS receiver in question was an old 6 channel one. There is nothing wrong
with 6 channels receivers, as long as their firmware propperly handles all
sats. This particular receiver did not do it. Trouble is, their full network
was the same receiver and viewing the GPS receiver as a risk did not seem to
have been on their radar. They had the rugg pull away under their feets.

Hopefully no major failure will result from enabling PRN32. But we can't be
sure. It's not like we are testing all our receivers in GPS simulators for
these kind of situations

Cheers,
Magnus