[sgreenhill]

I'm starting on a position-tracking application using the Nokia 6110. There's a trade-off between location sampling rate and battery life, but I can't find any official data on this relationship which is pretty central to this application.

The documentation on S60 Power and Resource Management says:

Using a location service has quite a dramatic impact on the battery life of a mobile device, especially when an integrated GPS positioning module is used. Developers must take power consumption into consideration when using this API. The general rule is to make as few location requests as possible. When tracking is needed, always select the largest update interval value that is acceptable. If a location fix is not required for a long time (that is, more than 10 minutes), the application should close the sub-session to the Location Server.

Presumably if I close the sub-session that allows the location server to power down the GPS radio and the power savings offset the costs of having to reacquire a fix on power-up. But short of doing this, how much power do I save by reducing the sampling rate? What sort of battery life can I expect from sampling rates of 1 second, 10 seconds, 60 seconds, 5 minutes, etc?

Does anyone have figures they can share?

[nicholso]

Using a Nokia E90 with an old battery, with J2ME application with a JSR-179 API location listener setLocationListener() set to 2 second update, the phone powered down after nearly ten hours of GPS usage. During this time the phone had the GSM/3G powered on to allow A-GPS but was in power standby (e.g.: display had clock screensaver).

[nicholso]

.. and when J2ME in current Nokia implementation (e.g.: S60 FP1) "close(s) the sub-session to location server", by calling a location reset during a GPS satellite fix, the GPS hardware does not power down. The only reliable solution to power down GPS is to exit the midlet.
http://wiki.forum.nokia.com/index.ph...d_battery_life