How exactly do GPS devices work?

[Muse]

I have a golf GPS, a Callaway uPro. When I got it a couple years ago I ruminated on how it worked and made some conjectures. I don't remember exactly, but I figured that my golf GPS has to send a signal, which probably includes a unique ID to one or more satellites, and the position of the GPS is calculated and it gets informed from the satellite concerning it's unique position, in this case distance from the front/middle/back of green, and possibly some other details based on the course and hole being played. Am I basically right about this? Would there be several satellites involved?

The golf rangefinders seem to use Lithium batteries whereas the GPS units (at least some of them) use Li-ion rechargeables. I figure that this is because the GPS units, having to communicate with satellites need more power, which is covered by frequent charges, whereas the rangefinders, which only have to bounce laser beams off close-by objects are not so power hungry, and the designers opt for the ease of Lithium batteries (only occasional replacement). I'd prefer a couple of Eneloop 1.5v but the laser rangefinder I just bought uses 3v CR2 Lithiums.

 

[MixMasterTang]

Sort of like you said but reverse, multiple satellites are sending out signals with info on where they are and what time the message was sent, you need signals from at least 3, and based on some algorithm your device tells you where you are.

 

[Muse]

Sort of like you said but reverse, multiple satellites are sending out signals with info on where they are and what time the message was sent, you need signals from at least 3, and based on some algorithm your device tells you where you are.

So, the GPS calculates its position based on the amount of time the signals took to reach it, using the speed of light in the equation? Wow!

 

[MixMasterTang]

So, the GPS calculates its position based on the amount of time the signals took to reach it, using the speed of light in the equation? Wow!

http://www.kowoma.de/en/gps/signals_runtime.htm

 

[Drako]

Triangulation

 

[Eureka]

It's intense how well the GPS system works simply by calculating location based off of satellites. The satellites have to be kept well calibrated, of course, and you even have to account for relativity.

 

[Muse]

Triangulation

Trilateration

So yeah, the GPS has to calculate its distance from at least 3 satellites and to do so it has to figure out how long it took the satellites' signals to reach it. The calculation involves the speed of light.

 

[dainthomas]

It's intense how well the GPS system works simply by calculating location based off of satellites. The satellites have to be kept well calibrated, of course, and you even have to account for relativity.

IIRC, there are ground stations around the globe that regularly send time corrections to the satellites.

 

[Jeff7]

They're fancy atomic clocks. In space.

Trilateration

So yeah, the GPS has to calculate its distance from at least 3 satellites and to do so it has to figure out how long it took the satellites' signals to reach it. The calculation involves the speed of light.

Even more fun: It involves the speed of time - relativity.

They're far away from Earth's gravity well, so that would tend to make their clocks run slightly faster than Earth-based clocks.
But they're also moving quickly, which would tend to make their clocks run slightly slower than Earth-based clocks.
They don't cancel out perfectly; they still need to account for the slight discrepancy caused by relativistic effects. (The GPS clocks end up running very slightly faster than Earth-based clocks.)

If they wouldn't account for that drift, your GPS directions would be very far off. Hundreds of feet, after only a day of drift.
<Apple Maps joke>


Science: Hell yeah.

 

[lupi]

Assuming there is still some error factor introduce to make them not as accurate as a military unit.

 

[slag]

Is it illegal to scramble GPS satellite data or attempt to cancel it out or hijack it in the US? I imagine it would be, but am curious.

 

[Fritzo]

GPS's work the same way as magnets.

 

[pete6032]

Is it illegal to scramble GPS satellite data or attempt to cancel it out or hijack it in the US? I imagine it would be, but am curious.

Yes. Your GPS signal and data is likely coming from a US government satellite. In the past during wars before GPS was a public thing, the US government would change satellite signal times so that the GPS would read incorrectly unless you had the correct offset information (at least this is what I was told in my GIS classes).

Here is an article about the evolution of GPS and wars.

http://abcnews.go.com/Business/story?id=86656

 

[el-Capitan]

Assuming there is still some error factor introduce to make them not as accurate as a military unit.

IIRC from navy days the time stamp that the satellites send can be offset for a particular region and altering the output of all GPS devices who are not fed with a particular (daily) code.

It is also more accurate. I think we had an additional decimal in our coordinates.

 

[Mark R]

Trilateration

So yeah, the GPS has to calculate its distance from at least 3 satellites and to do so it has to figure out how long it took the satellites' signals to reach it. The calculation involves the speed of light.

Actually 4 satellites. The GPS receiver does not know the range to each satellite. It only knows the pseudorange. In other words, it knows that sat B is 100.2 miles further away than Sat A, and that sat C is 242.1 miles further than Sat A, but it doesn't know how far away Sat A is.

3 pseudoranges is not enough information to compute the receiver's position in 3D coordinates. You need a minimum of 4 satellites in view to compute a 3D position.

Many receivers will attempt a crude calculation with 3 satellite signals locked. They will just assume that the receiver is at sea level and perform the calculations on that assumption. Because of the geometry, this isn't actually that bad an assumption, so on the surface of the earth, that a 3 satellite position estimate is rarely more than 100 yards away from the true position.

A series of ground stations constantly track the satellites (to within an inch or so) and calculate the details of their orbit. This orbital information, together with a reference atomic time signal is uploaded to the satellites. Each satellite beams down its own atomic time reference signal and it's orbital coordinates.

The receiver then downloads the satellite orbit parameters, and uses that and the time, to calculate the exact location of the satellites at any given time. Then from the multiple time signals from all satellites in view, it calculates the pseudoranges. With the pseudoranges and satellite positions, it is possible to calculate the position of the receiver.

The orbital "almanac" file is only transmitted by the satellites very slowly (about 1 bit per second), so it takes about 20 minutes to fully download. This is often inconvenient, because without an up-to-date almanac, a receiver can't compute locations (because the satellite orbits do drift a little bit over time). Modern GPS receivers (especially smart phones) will often use a wifi or cell data connection to download the almanac from a web site, instead of waiting for it download on the satellite signal. For E911, one solution that has been used, is for the phone to transmit the raw pseudorange data only. The 911 service has software with almanac pre-downloaded, which can then compute the phone's position with the pseudorange data and the cell network's atomic time signal.

 

[Coalfax]

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.

 

[Drako]

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.

You would be mistaken.

 

[Fritzo]

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.

I don't think they can fall out of orbit. They're quite a ways out there. The space station orbits around 250 miles up, where GPS satellites orbit at over 12000 miles up. If I remember correctly we need 24 satellites to keep GPS service operational, and we have over 30 in service right now, with a couple more scheduled to launch soon.

 

[Mark R]

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.

Europe has its own system. It has been tested with 4 satellites. The plan is to make the system functional by the end of next year, and completed with a total of 30 satellites by 2015.

The Galileo system, as it is called, boasts a large number of enhancements over GPS.

Stronger more error resistant signals; higher precision satellite clocks; direct correction of atmospheric distortion (speed of light through the atmosphere can vary with weather and solar conditions); more satellites in orbits that provide better coverage at extreme latitudes (GPS works poorly in the arctic circle and antarctica, as GPS satellites tend to hug the equator and tropics, so aren't visible at extreme North/South locations)

The signals are very similar to GPS, and the latest GPS receivers are (theoretically) upgradeable with a firmware update. It will also be possible for a single receiver to use both signals simultaneously for maximum reliability and highest precision.

It will be free to use. It will also feature an upgraded version of the GPS's military precision signal, which will be encrypted but civilians will be able to purchase a subscription to it. It is expected that the premium signal will offer precision in the 1-2 foot range.

 

[Ferzerp]

Assuming there is still some error factor introduce to make them not as accurate as a military unit.

No, not since 2000.

That's why consumer GPS Nav devices took off. It was nearly useless for anyone else while SA was on.


Also, they can't directly tell how long the signal takes to reach them (That would require clocks in each GPS nav device as accurate as the ones on the satellites, making them far too expensive). They can only tell the difference in the times they receive from the different satellites. This data can then be used to calculate the distances, which then gives the position.

 

[DesiPower]

It connects to the NSA servers and determines your location

 

[Mark R]

No, not since 2000.

That's why consumer GPS Nav devices took off. It was nearly useless for anyone else while SA was on.

There is no deliberate corruption of the civilian signal since 2000.

However, the civilian and military signals are different. The civilian signal is stronger, but less precise (it is formally called the coarse acquisition or C/A signal). The military signal offers substantially greater precision (it is the precise acquisition or P/A signal), but the signal is much fainter and much more complex requiring substantially more processing for the receiver to get a "lock".

In general, it is not computationally feasible to lock the military signal without assistance from first locking the civilian signal and using that as the starting point for a signal search. In addition, the military signal is encrypted, and without the encryption keys, the signal is virtually useless.

However, the real benefit of the military signal, is that it allows correction of atmospheric distortion. The atmosphere is not perfectly transparent to the GPS signals, and they get delayed and stretched. This delay and stretch is a major factor in GPS inaccuracy.

If a receiver can lock 2 signals simultaneously, and those signals respond to atmospheric distortion in different ways (the C/A and P/A signals do, and this is a major reason why they were chosen), then it becomes possible for the receiver to precisely calculate the amount of distortion, and subtract it out.

 

[_Rick_]

Europe has its own system. It has been tested with 4 satellites. The plan is to make the system functional by the end of next year, and completed with a total of 30 satellites by 2015.

The Galileo system, as it is called, boasts a large number of enhancements over GPS.

Stronger more error resistant signals; higher precision satellite clocks; direct correction of atmospheric distortion (speed of light through the atmosphere can vary with weather and solar conditions); more satellites in orbits that provide better coverage at extreme latitudes (GPS works poorly in the arctic circle and antarctica, as GPS satellites tend to hug the equator and tropics, so aren't visible at extreme North/South locations)

The signals are very similar to GPS, and the latest GPS receivers are (theoretically) upgradeable with a firmware update. It will also be possible for a single receiver to use both signals simultaneously for maximum reliability and highest precision.

It will be free to use. It will also feature an upgraded version of the GPS's military precision signal, which will be encrypted but civilians will be able to purchase a subscription to it. It is expected that the premium signal will offer precision in the 1-2 foot range.

Galileo and GPS (3?) will actually share a frequency band, if I am not mistaken, augmenting each other. Furthermore GLONASS and Beidou 2 are in development, with several recent setbacks for GLONASS, which is always teetering close to or in degraded mode, and there were I think two failed launches recently (2011-2013).

In addition to that there are also WAAS and local dGPS Systems to account for local/regional errors (e.g. what S.A. looked like, or timing drifts in the GPS birds, constellation errors, etc) and increase precision.

Fun fact: the previous generation of space localization platforms used the Doppler-effect instead of time-of-flight to localize a client device.

 

[Ferzerp]

No, the precision is exactly the same (and would have to be for this next sentence to have any meaning or validity)

You do get the ability to account for distortion with the two separate signals from the same location (as you mentioned), but that would be of limited use for us day to day.

What the encrypted signal does allow though, is the ability to be sure it isn't a spoofed signal.

 

[Mark R]

No, the precision is exactly the same (and would have to be for this next sentence to have any meaning or validity)

You do get the ability to account for distortion with the two separate signals from the same location (as you mentioned), but that would be of limited use for us day to day.

What the encrypted signal does allow though, is the ability to be sure it isn't a spoofed signal.

No. The P/A signal does have higher precision. One limit to precision is determination of the phase shift of the signal (because it is phase shift that is the measure of pseudorange).

The P/A signal has a much higher chip rate, this directly translates into a higher precision pseudorange for the same level of phase noise. (Which in practice is achievable due to the very high correlation gain of the P/A signal).

You are right about the anti-spoofing though. That is an important part of the P/A signal, which is notably missing from basic GPS.

This is another advancement of the Galileo system. In adition to the basic and premium signals, it will transmit a separate digitally signed signal which can be used for authenticity verification (although apparently only restricted sectors will be able to get access to the keys - e.g. aerospace, government, law enforcement).