user location

hi, anyone please -how precise can a gsm/umts operator using for
example a silent sms, locate a mobile phone user (with say a first
generation gsm phone without any special service software in it)?

---

› See More: user location

On 2008-01-04, [email protected] <[email protected]> wrote:
> hi, anyone please -how precise can a gsm/umts operator using for
> example a silent sms, locate a mobile phone user (with say a first
> generation gsm phone without any special service software in it)?

I don't know about the "silent sms" thing. You might be able to
find out the Cell ID from the phone that way, but the operator will
know this without the phone's help anyway. The Cell ID of the tower
the phone is attached to is the about only thing a vanilla GSM network
will know about the phone's location in any case, and the accuracy of
that will be related to the size of a cell; kilometers in the country,
maybe many 100's of meters in the city.

If the GSM operator has installed additional equipment at the towers
explicitly to do positioning, however, they can do better that.
This web page lists the stuff that can be done with additional
equipment:

http://www.trueposition.com/positioning.php

(note that A-GPS requires special hardware/software on the phone, but
everything else works with any GSM phone). U-TDOA is listed as supporting
sub-50 meter accuracy, though that is a best-case situation. The USA
E911 mandate requires that the operators be able to locate a phone to
within 100 meters 67% of the time, and within 300 meters 95% of the
time, and I've heard the U-TDOA system the biggest GSM operators are
installing is only barely capable of reaching this reliability.

Whether your GSM operator has installed additional equipment for
location determination is something I have no knowledge of, though.

Dennis Ferguson

[email protected] wrote:

> hi, anyone please -how precise can a gsm/umts operator using
> for example a silent sms, locate a mobile phone user (with
> say a first generation gsm phone without any special service
> software in it)?

It varies, and is not terribly precise.

Until there's an interaction with the network, the operator
knows only the cellular location area code (LAC) in which the
phone is presently located, the cell ID with which it last
communicated, and (potentially) the network measurement results
(NMR) and timing advance (TA) from that interaction.

Forcing an interaction updates that cell ID, NMR and TA
information so that the current position could be inferred.

TA is calculated by the network and passed to the phone, while
NMR is calculated by the phone and passed to the network.

Knowing TA permits the distance from the serving cell to be
estimated to an accuracy of about 550 metres.

NMR includes the signal strength measurements from the six
strongest neighbouring results. On flat featureless terrain,
this could permit position to be inferred with reasonable
accuracy. But I don't have a feeling for the degree of
accuracy in practice.

Note that TA is not a feature of UMTS networks.

John

I wrote:

> NMR includes the signal strength measurements from the six
> strongest neighbouring results.

That should read "neighbouring cells".

John