I've written a PyS60 application to log the position of handsets using the 'Network based' positioning method. I've been running this application on a Nokia E51 and a Nokia E71, both running the same version of PyS60 (1.4.5), and the accuracy readings that I'm receiving from the handsets are dramatically different.
The documentation I've managed to find suggests that the accuracy figure returned by positioning.position() is measured in metres, and the data from the E71 seems to support that. The cells are about the right size given the type of cell (2G/3G) and the geographic conditions (larger cells in flat or coastal areas).
The numbers I'm getting from the E51, though, are far too small to be realistic.
Here is a sample of the numbers I'm getting, where both handsets are connected to the same cell:
E51 accuracy: 89
E71 accuracy: 2470
E51 accuracy: 98
E71 accuracy: 4346
E51 accuracy: 102
E71 accuracy: 5770
E51 accuracy: 113
E71 accuracy: 11192
Not only is there a huge difference between the numbers, but there doesn't seem to be a linear relation between them. The equation y=(11x/200)^5, where x is the E51 figure and y is the E71 figure, seems to give a (very) rough approximation of the correlation.
So my questions are, why are the numbers different? If it is deliberate, then what units are the numbers in? Is there any way to get the E51 to report in metres, as the E71 seems to do?