Hello,

I'm using the GPS-API of the Mobile In The Hand Version 3.2 on a pda with
Windows Mobile 5.0. Mostly everything works quite fine, but from time to time
there are big slips in the gps coordinates and/or the time.

To be sure that the problem is not caused by anything from the bigger
application where the gps functions are intergrated I started a small project
with just the gps functions. But there are the same problems.

The project tries to store the actual gps position into a database every
minute. To preserve the resources the device I am only creating the GPS-Object
and listen to the events, when needed. So I create the GPS-Object every 60
seconds. The position_changed event will write the actual position into a global
variable and to be sure that the event would have been fired I wait a short
while. After this I will store the GPS-Data into the database and dispose the
gps object.

Analyzing the stored data I found out that something strange happened when slips
(sometimes more than 100 kilometres from the actual position) appeared. The date
belonging to the gps object was several minutes in the past. Even older than
previously stored data. After 5 minutes everything was fine again. And I got
correct data again.

Excerpt from a log (device time ! gps data)
--------------------------------------------
23:45:19: GPS-Data!13.09.08 21:44:51 53,0802016666667
23:46:21: GPS-Data!13.09.08 21:45:54 53,0802016666667
23:47:27: GPS-Data!13.09.08 21:47:00 53,079945
23:48:30: GPS-Data not valid
23:49:36: GPS-Data!13.09.08 21:44:46 53,0173466666667
23:50:38: GPS-Data not valid
23:51:45: GPS-Data!13.09.08 21:46:54 52,9989233333333
23:52:47: GPS-Data not valid
23:53:51: GPS-Data!13.09.08 21:49:01 52,9803716666667
23:54:53: GPS-Data not valid
23:56:10: GPS-Data!13.09.08 21:55:42 53,0803816666667

The slips occured at 23:51:45 and 23:53:51 and the date from the gps-data lies
in the past. The last correct data before war from 21:47.

Also sometimes I am getting gps data which have dates several years into the
future.

I would like to know if anyone else experiences such slips, if they result from
my "weird" method to get the data or if that might be some flaw in the gps
hardware (the same problem occurs on different devices of the same model).

Thanks in advance,
Anja

The problem you may have here is that if you are stopping and restarting the GPS device you have to allow for the time taken to get an accurate fix, it's quite normal in most cases for the first few readings to jump around until settling on a position. You can use the Horizontal/VerticalDilutionOfPrecision and Satellites.UsedInFixCount to determine the potential accuracy of a reading. Also the timestamp from the GPS is in UTC and so you should also log the UTC time in your log to compare like with like.

Peter