GPS Rollover

greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread

Could someone please provide a short and sweet summary of the GPS Rollover problem? I have seen it referenced numerous times and am unclear of the nature of the problem. Thanks.

-- bhayes (bkhayes@intellex.com), September 18, 1998

Answers

You can find an article on the GPS rollover at: http://www.year2000.com/y2karticles.html

-- John Callon (jcallon@gate.net), September 18, 1998.

As I understand it, the clock chip in the Global Positioning Satellites has a capacity of 1024 weeks. On rollover date (I think it is August 22, 1999), the clock reaches its maximum and rolls over to 0000 again. GPS receivers use some mathematics to compare the value of the time signal, taking into account the time-delay caused by distance the signal travels, from multiple GPS satellites to determine their distance from each of these satellites and thus the position, altitude, etc of the receiver. Some GPS receivers (older ones) apparently may be confused by the signals when they roll over to 0000. It is claimed that newer devices have been made "Rollover Compliant".

It is also my understanding that digital communications transmissions require very precise time signals (at the nanosecond level?) to operate properly (such things as digital cellular phones). The communications companies have latched onto the extreme precision of the GPS time transmitters and built this into their equipment. The concern expressed is that there is risk that certain technologies, dependent on the GPS timing signals, may fail at GPS Rollover time.

The whole situation has many parallels to Y2k, but occurs a few months earlier.

<<<<<<<<<<<<>>>>>>>>>>

-- Dan Hunt (dhunt@hostscorp.com), September 18, 1998.


Also note, on Ed Yourdon's homepage, in the left hand column, there is a click through to GPS satellite sytem, in which he has an excellent little chart available. Today it is saying "Only 338 days to Aug 22, 1999 when GPS satellite system rolls over." <<<<<<<

-- Dan Hunt (dhunt@hostscorp.com), September 18, 1998.

Global Positioning System satellite network ticking toward problems

09/18/98 05:20:13 AM By John Diedrich

The Gazette (KRT)

When the government turned on the Global Positioning System satellite network in 1980, a nearly 20-year clock started ticking.

At exactly 6:59 p.m. Eastern Daylight Time on Aug. 21, 1999, the clock starts over at zero. And that could cause problems for some commercial GPS users.

GPS began as a tool to help the military. A ring of orbiting satellites shoot back readings to receivers, giving location in latitude and, longitude. Just like the Internet, GPS now is used by millions, including farmers, truck drivers and surveyors.

Engineers programmed GPS satellites to run for 1,024 weeks or just over 19.5 years because of limited memory in the system's internal clock. When the time's up, the internal clocks start over for another 1,024 weeks.

The rollover won't have an impact on the 24 GPS satellites, which are operated from Schriever Air Force Base, or the government's receivers.

But some of the first generation commercial receivers built in the early 1990s might be affected. They weren't built with the rollover in mind and could freeze up or give slightly faulty readings.

Users should contact the company that manufactured their unit to check and get a possible fix.

Another problem facing GPS users is the millennium bug, which is expected to bite the world's computer systems on Jan. 1, 2000.

If the GPS receiver depends on an internal clock, the millennium change could freeze the unit. Just as importantly, GPS receivers might appear to malfunction if the computers they are hooked up to aren't fixed.

Officials in charge of GPS are anxious to notify the public about the possible impending problems.

``So many rely on GPS, we want to make sure people are aware of them,'' said Aaron Renenger, a spokesman for the military agency that manages the system.

X X X

(c) 1998, The Gazette (Colorado Springs, Colo.).

http://web3.stlnet.com/postnet/news/wires.nsf/National/E637E79D87DB7C588625668300380109

-- Pastor Chris (pastorchris@lifetel.com), September 18, 1998.


the U.S. Naval Observatory has a page about the GPS rollover at http://tycho.usno.navy.mil/gps_week.html

and the Air Force has a GPS site with links galore at http://www.laafb.af.mil/SMC/CZ/homepage/

-- John Howard, Greenville, NC (pcdir@prodigy.net), September 18, 1998.



I have had to deal with this one to some extent, and it seems the older receivers are just going to give bad data. Now the good news. THe newer receivers are MUCH cheaper than the old ones, give better service, and don't have the rollover problem. The Corps uses a lot of these things, and intends to replace most of them in the next year of so. Not a major problem for us, anyway, when you operate on the Mississippi River you pretty well can stop and ask where you are anyway. It wouldn't be critical unless you had a flood at just that time - very unlikely to have a flood in January. March is still pretty early for spring flooding, and the dredges stay in port for overhaul until Spring.

-- Paul Davis (davisp1953@yahoo.com), September 19, 1998.

Paul, you lost me at this point:

"It wouldn't be critical unless you had a flood at just that time - very unlikely to have a flood in January. March is still pretty early for spring flooding, and the dredges stay in port for overhaul until Spring."

GPS rollover is August 22nd, 1999. Did I miss something?

-- Gayla Dunbar (privacy@please.com), September 19, 1998.


I meant the positioning data is nice to have but not really necessary unless we have a flood. You just don't flood in the Mississippi Basin in the fall.

-- Paul Davis (davisp1953@yahoo.com), September 19, 1998.

Sorry I cut that short - got interrupted. The dredges come in in the fall for overhaul. (also I must admit I still sometimes get stuck on the 1/1/00 date) While they are in they don't need positioning date. The guys who might have trouble then are the ones with the newer receivers, marking levees for survey and such, who should not have a problem. The older ones are scheduled for replacement. Besides, the main use we have for them is to locate a precise point to start dredging the channel to keep it open for barge traffic - nice but not a really critical thing - if you dredge a bit more than was really necessary you wasted some money, but the job still got done. Its not like we were 1000 miles from nowhere and had no idea of our position within 50 miles. That gets you into trouble really fast with collisions and such getting really possible. At least on the Mississippi you know where you are and where you are going.

-- Paul Davis (davisp1953@yahoo.com), September 19, 1998.

I was under the impression that the problem was with the GPS units that used the satellites for time synchronization, since they have a very accurate clock. I know very little about GPS - Are the positioning and time functions related? In other words: Would a unit with the rollover problem be unable to track position properly?

On a related note - I know some clocks update themselves based on signals from atomic clocks operated by the government. There is a very sophisticated signaling system via radio from Ft. Collins, Colorado and Kekaha, Hawaii. I haven't heard anything about problems in this area. Is anyone aware of any glitches this might cause?

-- Mike (gartner@execpc.com), September 19, 1998.



Technically: the entire thing with the satellites is timing and remaining in "synch": the receivers use the difference in time between the signals + doppler effect from several different satellites in the network to calculate the receiver's position with respect to the moving satellites. (One result of the calulations over time is the speed and elevation of the receiver, for example, as well as its x,y position)

That's why you need to wait several minutes to get your "first" position, and why, before the latest extra satellites were launched you would have to wait until your reciever could track three or four satellites.

The satellites have to remain in synch witheach other and the ground: that's part of the broadcast signal that other programmers are using to "piggyback" data from for other uses.

-- Robert A. Cook, P.E. (Kennesaw, GA) (cook.r@csaatl.com), September 20, 1998.


Robert:

If that's the case, then I could see some positioning problems at the turn of the millenium. Say a non-compliant unit was trying to do a calculation and coming up with a negative number. But does this apply once the rollover is complete? I would think that as soon as all of the signals are back to month #1, (or zero, however they count) then the calculations would not pose a problem. I'm still scratching my head! (I should probably look around for more info about how the GPS works.)

BTW - Maybe this relates to the "Why me?" thread. CURIOSITY and a need to satisfy it, either through knowing what will happen, or being prepared for what might.

-- Mike (gartner@execpc.com), September 20, 1998.


Your questions are the very ones "they" don't always know the answers to; plus, the answers will be different for each different receiver and application, and the solutions may or may not have been applied correctly in every different application.

That level of complexity in just finding out the "questions to be asked" is why the inherent threat of Y2K in general to the existing nest of interrelated systems isn't understood by very many people.

You've made a great start by looking for answers.

Congratulations, but the final exam in Y2k classes is: answer all the questions in the following list completely and honestly. You will be graded on a pass/fail basis. Every question must be answered in complete sentences and be referenced in the bibliography. Failure means extreme discomfort, possibly illness, poverty or death.

You get to define the questions.

And the rest of the exam paper is blank.

-- Robert A. Cook, P.E. (Kennesaw, GA) (cook.r@csaatl.com), September 20, 1998.


Moderation questions? read the FAQ