2

I hope I'm asking this at the right part of Stack Exchange. Please bear with me if I'm wrong.

I'm developing some gps based applications. The demands I have for precision are not very high, but I need to know the possible errors.

I have learned that the fourth decimal gives you ~10 meters precision. That should be enough for me.

The real question is how fast will I get that precision realiably in different environments (indoors, outdoors free sky, forest, cloudy, city etc).

The applications I'm developing is for handheld devices so I prefer to have the gps active in as short intervals as possible.

As I do now, the intervals I use the gps are more governed by battery life than precision. Now I'm trying to balance the two.

Community
  • 1
Einar Sundgren
  • 123
  • 4
  • 1
    What about cold start and warm starts to hardware? http://www.maptoaster.com/maptoaster-topo-nz/articles/how-gps-works/how-gps-works.html – Mapperz Nov 01 '13 at 19:13
  • Good link @Mapperz. Thanks. I world say I most of the time will be using hot or warm starts to the hardware. – Einar Sundgren Nov 01 '13 at 22:04

1 Answers1

3

There is no meaningful way to link "time the GPS is on" with a particular precision level. The accuracy is driven by many factors ranging from the characteristics of the hardware (especially the antenna) through to atmospheric conditions and GPS satellite constellation configuration.

A reasonable approach (using information you can get from most receivers) is Dilution of Precision. If you only care about 2D positioning, the best indicator of accuracy is HDOP. If you also care about altitude, the best indicator of accuracy is PDOP.

So instead of trying for a fixed time, monitor DOP.

BradHards
  • 12,881
  • 2
  • 37
  • 70
  • From other reading I have done I get that you are right. I was actually not aware of this DOP property of that most receivers give you. But that solves my problem in an easier and more reliable way than using approximate times. – Einar Sundgren Nov 02 '13 at 10:48