GPS accuracy with software

Question

I spoke with a guy who claimed that he's working (programming) on a GPS-system that is going to be accurate to the centimeter.

• Is this even possible?
• What is the limiting factor, the big distance between the satellite and the GPS, or is it the software?
• Can the software be improved to make a GPS accurate to centimeters?

Answer 1

In the good old days the US DoD had a thing called "Selective Avalibilty" -SA- - they turned this off at the beginning of the century making Civilain use of GPS more accurate, enabling things like car navigation etc. Prior to that Differenctial GPS was a widely used system for those that required greater accuracy than could be achieved with SA- DGPS is done using a ground base station at a known coordinate, that transmits the error in the GPS signal to close by GPS recievers.

Using DGPS is still possible to increase the accuracy of a GPS fix today, although its rarely used as GPS is currently accurate enough for most purposes. Other means to get higher accuracy are to plonk the receiver down for anything from an hour to a day or two and average the fix. (Surveyors use this technique.) - you need to understand what GPS accuracy is - if they say it's accurate to 10 meters, they usually mean "you are with in 10 meteres 90% of the time" or something like that. I have never seen a GPS state a 100% of the time figure. It is therefore entirely possible to get sub-centimeter fixes from GPS (e.g. A marketing guy might claim 1 cm and then in the fine print put "Accurate to 10mm 10% of the time".

Why is this accuracy needed in your friends case? There are a few cases (surveying) where it is. The earths model (WGS84) used by GPS is not that accruate, let alone the maps, and unlikely to be for along time, whats the point of being out by less than 1 centimetre when the error in your map has more than 1 meter error?

Answer 2

It is my understanding that by gathering data over time you can get that kind of accuracy out of GPS. It requires a unit that is not moving, it's for survey work and the like, not navigation.

Answer 3

There is a limit for averaging GPS position due to real random variation in the atmosphere. With single channel L1 civilian GPS this is typically 10-20 cm after 10-20mins. Military GPS uses a second channel L2 at a slightly difference frequency so it can estimate some of the atmospheric effect and allow for it.

It is possible to make cm accurate differential readings of the relative positions of two nearby receivers over the same time period - since the atmosphere effects will cancel out - although you won't know your absolute position on Earth any more accurately. You can get a differential signal from a broadcast fixed beacon in a known place then you have an accurate and known absolute position. There are networks of fixed beacons run either by national mapping agencies or GPS makers - and depending how far away the nearest one is - tis can also give cm precision in a few minutes.

With special survey units which can communicate and transmit the signal phase (Real Time Kinematic - RTK GPS) you can make mm accurate position measurements of a moving GPS unit in real time, relative to a fixed base station.

Answer 4

Civilan GPS signals are intentionally degraded because they only use one signal. All the details are available online at GPS.gov.

Military GPS uses two singles to be much more accurate.

Civilian devices can be augmented with other signal sources to be more accurate than the published Military specs. Of course the Military can augment their systems even more the same way.

Is Military GPS More Accurate Than Civilian GPS?

The accuracy of the GPS signal in space is actually the same for both the civilian GPS service (SPS) and the military GPS service (PPS). However, SPS broadcasts on only one frequency, while PPS uses two. This means military users can perform ionospheric correction, a technique that reduces radio degradation caused by the Earth's atmosphere. With less degradation, PPS provides better accuracy than the basic SPS.

It takes more hardware to make the device more accurate, more sensors reading more signals. Without the augmented signals, there is nothing that software alone can do that isn't already being done.

Answer 5

Carrier phase tracking can achieve an accuracy of 0.2 centimeter. At that point, you need to clearly define "accuracy" because maps, or even the earth's shape aren't that well defined.

Answer 6

You can use software to compare the readings from 2 (or more) matched GPS receivers which are looking at the same set of satellites to cancel out any differences in (the many sources of) absolute errors. If you stick one on a centimeter accurate benchmark, and the other is nearby with the same sky view, you can use software to compute the difference and add that to the known benchmark location.

Answer 7

Modern receivers use A-GPS (Assisted GPS), which calculates location based on:

• US DoD's GPS
• GLONASS
• (soon Galileo and Compass)
• WiFi hotspots
• 3G base stations

It's not just averaging results from various sources, but actually using all of them to create better results.

Also, if the receiver remains stationary for longer it's possible to increase accuracy a lot. This is the method used typically in surveys.