Java Development

[Pascal_Bourque]

Hi all,

   Here my problem, i wrote a program to geolocation BB phone on road and calculate de miles driven. To do this I use a location provider that send me a position, at second, and a calculate de distance between each position. It work perfectly except that after 7 hours ago I have no battery. It worst if the provider fail to find a position because I reset the provider immediatly and restart the provider.

   I need acurate position a quick as possible(TTFF) and have an high ratio of position to calcul the distance, also I need the battery stay alive for 12 hours, so what suggestion do you have ?

Thanks,

[Pascal_Bourque]

Here some more info,

   The JDE I use is 4.5.0 because I work on BB8830, Curve, Bold8900, Tour and Storm.

[RexDoug]

I don't see any point on "per second" fix. Why don't you try backing this interval down and see if that improves battery life?

I would try 5 or 10 seconds.

[klyubin]

Are you using a GPS-only provider (i.e., a provider that does not need to communicate with the cellular/mobile network) or a provider that talks to the cellular network to increase the accuracy of the fix? I suspect the latter will drain the battery quicker, although I'm not an expert of location providers.

[RexDoug]

In my experience autonomous GPS drains the battery quicker.

Network-assisted is not accurate enough for real-time navigation anyway.

[klyubin]

Interesting!

P.S. I actually had A-GPS in mind, or a provider that combines the network-based location with GPS-provided location to increase the precision (not sure if something like this would actually work).

[peter_strange]

At the DevCon 2008, they recommended 10 seconds or less as the best time period for getting GPS locations that minimized the drain but kept the chip hot.  Using that I'm lucky if I get a day out of a battery. let alone 7 days.

Also I've done a bit of work on the accuracy of GPS.  I would be very suspicious of any Tracking that accumulated the distance moved in every GPS reading that it took.  I'm sure you have all seen how your GPS location can move around even when you are stationary.  We attempted to overcome this problem by comparing current readings and previous ones and only accepting a new location when the distance seemed reasonable given the time period involved.  That works OK, better than nothing at all anyway.  But it is non trivial.