How exactly do GPS devices work?

[Muse]

The sad part of all of this is that the sats that are responsible for this are getting pretty old, and currently I don't believe that there are funds appropriated to replace them as they fall out of orbit.

Oh, you better believe that the funds will be found. GPS's are far too important (i.e. used by so many people who depend on them, sometimes love them).

Edit: Anyway, as noted, they are at around 12,000 orbits and those are not going to degrade. The link I made earlier indicates there are extra satellites up there, just in case.

 

[Drako]

Oh, you better believe that the funds will be found. GPS's are far too important (i.e. used by so many people who depend on them, sometimes love them).

Edit: Anyway, as noted, they are at around 12,000 orbits and those are not going to degrade. The link I made earlier indicates there are extra satellites up there, just in case.

LOL, he's clueless. I know people at Schriever, and they are ready.

Simple web searches would tell you so.

 

[JimmiG]

Triangulation and magic.

 

[Colt45]

I figure that this is because the GPS units, having to communicate with satellites need more power,

The GPS don't 'communicate' with the satellites as such, they only receive, which requires orders of magnitude less power than transmitting.

The unit you have might download maps over mobile networks or somesuch, though, which is kinda heavy on power.

 

[John Connor]

Don't forget WAAS. http://en.wikipedia.org/wiki/Waas

DARPA is developing a chip that has two gyros, a timer and accelerometer to augmentate GPS. This means your smart phones might not even need GPS and the battery will last longer. http://www.darpa.mil/Our_Work/MTO/P...ble_Gyro_for_Absolute_Reference_(COUGAR).aspx

 

[Muse]

Based on what I'm seeing here I figure GPS systems will get better and better, more and more accurate and dependable. One reason I just bought a line of sight laser rangefinder is my golf GPS often struggles to determine position. Its readings can be several yards off, sometimes more, occasionally a lot more. So, I can't use it with confidence. I'm told that the laser rangefinders are dead on. I don't know if my golf GPS will get better over time if problems are worked out with its satellites, or maybe its doomed to inconsistency. Maybe satellite improvements and firmware updates can work out some things.

The GPS don't 'communicate' with the satellites as such, they only receive, which requires orders of magnitude less power than transmitting.

The unit you have might download maps over mobile networks or somesuch, though, which is kinda heavy on power.

I'm pretty sure it doesn't use mobile networks. It does use maps, don't know how they implement that. I can choose from any of a couple of dozen courses I've bookmarked (or downloaded, don't know how it works). I choose the course after starting it up, but it takes a few minutes before it sees where it is and figures out what time it is.

 

[_Rick_]

Don't forget WAAS. http://en.wikipedia.org/wiki/Waas

DARPA is developing a chip that has two gyros, a timer and accelerometer to augmentate GPS. This means your smart phones might not even need GPS and the battery will last longer. http://www.darpa.mil/Our_Work/MTO/P...ble_Gyro_for_Absolute_Reference_(COUGAR).aspx

Everything I know about INS tells me, that this is not going to work.

Controlling drift even over minutes using hardware at a comparable cost to a GPS receiver is currently impossible. MEMS advances may improve the precision of "silicon"-INS, but even then, I don't believe in more than a few minutes of reasonable precision, after that drift will make the results worthless.
I can imagine it as a solution to roughly navigate indoors, but currently semantic approaches appear to be more promising (i.e. combining a floorplan with a podometer and using floorplan constraints to attempt localization)

 

[Imp]

No mention of the Russian GLONASS system? It's supposed to help with positioning and relatively new. Kills the battery though.

From what I've heard, survey-grade GPS receivers should be accurate to centimetres, so it's just a matter of the tech filtering down to consumer-grade stuff.

 

[skimple]

It connects to the NSA servers and determines your location

LOL!

 

[kevinsbane]

No mention of the Russian GLONASS system? It's supposed to help with positioning and relatively new. Kills the battery though.

From what I've heard, survey-grade GPS receivers should be accurate to centimetres, so it's just a matter of the tech filtering down to consumer-grade stuff.

Survey grade GPS units use a differential carrier-wave analysis, using "offline" ephemeris data.

Essentially, it requires two GPS units, and performs a differential analysis between the two receivers, to determine the exact number of waveforms in the signal. Thus it is not time-based in its measurement per se but rather resolves an ambiguity (the number of full wavelengths between the GPS satellite and the receiver). Having measured this, and having the receiver measure the fraction of the wavelength received, makes it possible for the receiver to determine there are, say, 150395.4490 wavelengths between the GPS satellite and the receiver. Multiply by the wavelength, and you get an exact distance, without having to involve time in your measurement. Mind you, timing is still essential, but it doesn't have as crucial a role in determining the receiver's distance to individual satellites. Earlier consumer models made use of the data encoded in the GPS signal to determine the time for a signal, which you multiply by the speed of light to give you a distance. This required very precise clocks, as the difference of a nanosecond will result in ~0.3m of error. That's not to say that you can have less accurate clocks in survey equipment though!

Survey equipment then has a couple of incidental advantages:
1. It is not reliant on the encoded GPS data, thus is much quicker and allows for real-time centimetre level accuracy even as you move around. Consumer GPS devices will "jump" around in a small radius, and may lag slightly as you move around. Survey equipment will not jump around. Once lock is achieved, location is precise, steady and real-time.
2. As it is not reliant on the encoded data, it can make use of the P/A signal as well as the C/A signal, since it only cares about the carrier wave itself, not the data that the carrier wave encodes; which allows for even greater accuracy and reliability. In fact, before selective availability was turned off, this technique (differential GPS) allowed for the use of the P/A signal to cancel out the effects of SA on the accuracy of GPS.

Of course, at that level of precision, many different factors start to affect your GPS accuracy - time of day, temperature, multipath (highly reflective surfaces nearby), solar weather, latitude, and even orientation of your GPS unit.

Edit: more sources of "error":
Geiod model, coordinate system (UTM versus a local system, NAD27 vs NAD83 vs WGS84), humidity, atmospheric composition, air pressure

All of those will have a significant effect on your centimetre-level accuracy.

 

[John Connor]

Everything I know about INS tells me, that this is not going to work.

Controlling drift even over minutes using hardware at a comparable cost to a GPS receiver is currently impossible. MEMS advances may improve the precision of "silicon"-INS, but even then, I don't believe in more than a few minutes of reasonable precision, after that drift will make the results worthless.
I can imagine it as a solution to roughly navigate indoors, but currently semantic approaches appear to be more promising (i.e. combining a floorplan with a podometer and using floorplan constraints to attempt localization)

Well, INS is primary how commercial airplanes fly.

 

[Mark R]

No mention of the Russian GLONASS system? It's supposed to help with positioning and relatively new. Kills the battery though.

From what I've heard, survey-grade GPS receivers should be accurate to centimetres, so it's just a matter of the tech filtering down to consumer-grade stuff.

GLONASS isn't new, it's been around for about 20 years, but has been so hopelessly unreliable, that it's basically been regarded as useless. The signals are broadly compatible with GPS, so making a dual-system receiver is fairly straightforward if you've already managed to make a GPS receiver.

For a large part of its life, the GLONASS system has been plagued with satellite failures, launch failures and depletion of funds. As a result, there are only small areas of the world where 4 satellites are visible at any one time.

The Russians have been trying to get the system fully-operational again recently with newly redesigned satellites, but there have been some issues. Most recently, a bulk launch of 3 satellites suffered technical difficulties because a greasemonkey installed the INS upside down on the rocket.

GLONASS satellite launch