[LEAPSECS] Problems with GPS?
magnus at rubidium.dyndns.org
Fri Feb 12 12:48:50 EST 2010
Tom Van Baak wrote:
>> So what I'm hearing is that the group is unaware of any specific
>> issues with civilian GPS units?
> In the past few months some civilian receivers experienced
> glitches that were eventually correlated to changes made in
> ground control software or something like that. The changes
> were still ICD compliant but apparently they exposed latent
> firmware bugs in a few gps receivers -- receivers that had
> otherwise been working flawlessly for more than a decade.
> Sorry, I don't have first-hand information on the problem.
> This sort of thing seems to happen every few years. Most
> recently recall that some receivers mishandled the 32 SV
> boundary case. Prior to that there were Y2K and WNRO
> fixes. Probably other cases as well; I don't have a full list.
Actually, the SVN31 issue caused more trouble than the SVN32 as I recall
it. Those where not changes in the ICD (actually it is, but it occurred
a long time ago) but in the use of the ICD. When the expected full
system was 24 SVNs, then including support for all 32 SVNs was not
obvious a wise choice. Firmwares can have many implicit deviation in
support which can be hidden for over 15 years before they bite. Vendors
could fail, be bought by others and drop support. Users should consider
those aspects to evaluate how long they can maintain GPS receivers in
operation and eventually decommission them from service when the needed
FW support/upgrades is not available any more and expected compliance to
ICD and its use among other things. Then again, if they judge that GPS
receivers have no impact on their operation then all is fine and they
can die safely... if only the rest of the environment is protected.
More information about the LEAPSECS