Satellite Glitch Rekindles GPS Concerns 101
coondoggie writes "News today that the Air Force is investigating signal problems with its latest Global Positioning System satellite is likely to rekindle the flames of a congressional report last month that said the current GPS coverage may not be so ubiquitous in the future.
The Air Force stated that routine early orbit checkout procedures determined that the signals from the Lockheed-built GPS IIR-2 (M), which was launched in March, were inconsistent with the performance of other GPS IIR-M satellites.
The Air Force said it has identified several parameters in the GPS IIR-20 (M)'s navigation message that can be corrected to bring the satellite into compliance with current GPS Performance Standards."
Re:Soloution? (Score:5, Informative)
Soloution? Pour more money into NASA!
Um, the GPS constellation belongs to the USAF.
Re:Followup on the story (Score:5, Informative)
Pretty much. Of course, any current problems with GPS are likely to be fixed once GPS III [globalsecurity.org] is up and running. It will combine land-based positioners with satellites that have 500 times the transmitter power. Of course, that 'not so ubiquitous' factor is still there as GPS III will allow the U.S. military to shut down GPS to selected geographic areas at will to all but sanctioned receivers. ;)
Re:GPS will be just fine (Score:5, Informative)
This is a little more serious than just some glitches in the software. It's a basic design problem.
Re:Followup on the story (Score:3, Informative)
Well, if you don't like it, you're free to build and deploy your own damned GPS system. Don't have hundreds of billions of dollars to spare? Don't have a national interest in precision, real-time mapping? Then shut the hell up.
You called? [upi.com]
Re:yet another justification (Score:3, Informative)
It's not that it doesn't work, it's just that if your GPS happens to be getting a time signal from that one particular satellite, the accuracy might be degraded. The article said +- 20 ft. That's not a big deal. It may be if you were using the GPS to land a plane though. A GPS only needs 3 time signals to triangulate, but can be more accurate if it can see more signals.
Re:What's The problem? (Score:3, Informative)
"... which was launched in March". That is not "getting on a bit" - it says that replacements are not fully-functioning copies of the originals, which is worrying.
"The Air Force said it has identified several parameters in the GPS IIR-20 (M)'s navigation message that can be corrected to bring the satellite into compliance with current GPS Performance Standards"
In other words, a small workaround is needed, but the satellite will be just fine. If you only knew how often this happened in engineering.
"The degraded signals are accurate only to about 20 feet, versus about two feet for typical GPS signals, the article stated."
Sure, this should have been picked up in testing, but it's not like testing something like this is easy. The accuracy needed to detect something of this magnitude is pretty staggering.
Re:Soloution? (Score:3, Informative)
Re:Soloution? (Score:3, Informative)
Re:solution: (Score:2, Informative)
I'm a man, I know exactly where I'm going and how to get there. If I get lost, I don't need a GPS nor anyone else to get back on track. If I don't get back on track, at least I've discovered new lands and people which is all part of the journey. My moto: delivery in more than 30 minutes or its free.
Re:yet another justification (Score:4, Informative)
Heck, I work for a GPS company and don't rely on GPS for navigation (we do high-precision applications, not navigation so I don't feel too hypocritical about it). However, this isn't because I don't trust GPS but because I prefer to have a good situational awareness of where I am and where I'm going.
I have tried using GPS navigation in rental cars for fun but usually their directions aren't as good as Google maps directions and have led me down a dead-end street once late at night.
The best compromise I've seen so far between maintaining a good idea of what's going on and using GPS is Google maps on the G1. You simply map out your directions before you start driving and then you have the choice of enabling the GPS to see where you are on the route or leaving it disabled if you like. Either way you can still zoom in on the map, check street view, etc without needing to spread out a huge map (or having to print out directions before you take off).
Note: I wouldn't use this story as a justification for not using GPS. The Air Force likes to maintain a 95% reliability of the constellation remaining fully operational each year (meaning that in the lower 48 that you will get a good position fix virtually all the time). They are worried that in the future they can only estimate an 80% chance of the system staying fully operational. The system would still work even if they don't have 31 satellites working. The minimum number of satellites needed in the constellation to provide good position fixes virtually all the time is 24 (4 good satellites in 6 orbital planes). They have additional satellites up there that are either at their end of life or backup satellites that are ready to take the position of another that becomes disabled. See this PDF if you want all the details: http://www.navcen.uscg.gov/GPS/geninfo/2008SPSPerformanceStandardFINAL.pdf [uscg.gov]
Re:yet another justification (Score:3, Informative)
A GPS only needs 3 time signals to triangulate,
it can continue to track with 3 time signals (since it can rule out some incorrect solutions, based on a physical movement constraint), but can only triangulate with a minimum of 4 inputs. 3 points works in a 2d (IE draw circles on a paper) sense, GPS are calculating intersections of 3d Orbs. It is also possible for terrestrial GPS to assume your the solution located on the surface of earth. Technically a GPS only needs some 4th bit of info, since you are always moving in relation to the satellites, it is possible to use that movement to fine tune your position over time, but most handheld GPS's don't have that bit of calculus and require at least 4 satellites, then assume a maximum movement if you drop to 3.
Re:What's The problem? (Score:4, Informative)
The satellites are not identical by any means. Rockwell built the originals and they were rock solid R&D birds. SVN3 was nearly 13 years old when I (and the crew I was on duty with) turned off the payload back in the early 90s. I used to joke with the Rockwell engineers that if requirements could be stopped, a Block 1 payload/chassis, with Block 2 electrical system would last decades (plural). Instead, we've got L1, L2, M-Code, L-5, NUDET territary sensors, and there's some boxes that only 3 letters now know about... Too much crap on something that's basically just transmitting, "HERE I AM!! (at this time)"
Re:yet another justification (Score:3, Informative)
A false solution way off the earth, but not all of them. If you imagine 2 spheres interacting where the satellites are both straight above you at very close distance, you end up with possible solutions being in one big elliptical arc, a good portion being away from the earth. When you add in a third source, they must intersect in at least 2 locations. But because the satellites are very close together your most likely at 4 intersection points, 2 will be in deep space, 2 will be in the earths sphere. The further apart the centers of the spheres the more circular the solution from 2 satellites, the less likely multiple solutions. So if you got good solutions from satellites on opposite ends of the earth, then 3 is good, you won't get that unless your really high though.
Of course that is all assuming a perfect measurement, add in a little uncertainty in the time dimension (all of your distances are going to be a slightly different moments in time, but within a few microSeconds) and with the closeness of the GPS satellites (relative to your location) your 4th source is a absolute requirement.
As you say, a little ground based info goes a long ways, partially because it is from such a different direction and distance, its accuracy can even be less.