Mobile phone localization accuracy

[Danijel Gorupec]

How precise can a mobile phone operator locate a hand set position? What factors contribute to that precision?

 

I was thinking if it is possible for a mobile phone operator to simply exclude a single building (or even large room inside a building) from the mobile-phone service? For example, would it be possible to arrange with mobile phone operators to exclude cinema halls or school classrooms from its service?

 

I am aware that mobile phone operators would have no interest in doing so, but law can be adjusted so that they must comply (in some cases, at least)... Do you already know a mobile phone operator offering such posiblitiy?

 

[swansont]

Location precision of the Uplink-Time Difference of Arrival protocol is about 50 meters.

http://en.wikipedia.org/wiki/U-TDOA

 

Not sure what the relevant number is for assisted-GPS.

 

You could probably exclude schools, suburban ones, at least as they are usually surrounded by some open area, but that renders using a cell phone for a 911 call moot.

[Danijel Gorupec]

If I understood the article correctly, the U-TDOA and GSM have approx. equal accuracy. Still, 50m error seems not good enough to isolate a room - 5m is needed.

 

I suppose that:

- if the error has no bias (low precision), then better results could be obtained by longer-time averaging.

- if the error has bias (low accuracy), then, maybe, it is possible to search for large groups of still-standing mobile phones (large concentration of mobile phones on small area) near the target room location and conclude that this actually is the target room.

 

(Note: not all service features need to be switched off, so 911 calls could remain)