Cell Site Analysis - GPRS

Can any cell site analysts here share their thoughts regarding this issue please…

I suspect no one has responded because you need to provide more info about your gprs tech case. Each operator handles this matter differently.

Thanks for the reply,

I suspect no one has responded because you need to provide more info about your gprs tech case.

I understand that there is nothing about the case in question but all I can say at this moment is that the network in question is H3G. But, I think the issue is irrespective of the case. This is applicable to all other networks as well.

There is no doubt that the reported Cell ID will definitely be utilised by the phone exactly at the date/time shown if it was used for a text or voice call. But unfortunately there are varying explanations regarding the cell use for GRPS usage.

Each operator handles this matter differently.

Exactly! I was hoping that someone with experience in this field could shed some light into how each operator records GPRS usage and most importantly what does the date/time associated with the cell usage signify.

Many thanks.

Thanks for your reply.

Reference these two statements

2. A second report by ACPO that, the cell id shown may be used by the phone at or before the time shown, but not necessarily at the exact date/time shown.

3. A third explanation states that, the CDR will only show the the 1st cell that it encounters in its routing area and as long as the phone stays in the same routing area, the CDR will show this cell for subsequent GPRS usage. If the phone changes the routing area, the reported Cell ID will the the 1st cell id that the phone encounters in the new routing area.

Can you confirm the precise documents in which these two statements are quoted?

Thanks

Thanks for the reply.

The 2nd Document is a privately distributed and "protected" document to cell site contractors. I assume that cell site contractors will have access to that.

The third explanation is the one given by most of the cell site analysts - there is no recorded document or published paper for that.

Thanks.

Thanks for the reply.

The 2nd Document is a privately distributed and "protected" document to cell site contractors. I assume that cell site contractors will have access to that.

The third explanation is the one given by most of the cell site analysts - there is no recorded document or published paper for that.

Thanks.

subujoseph, I am really grateful for the time and help you have given to this matter and I apologise asking so many questions. I am curious about the GPRS comments being linked to the third explanation. How many of those analysts that you mention say this can actually demonstrate a deep-level independent, objective and impartial assessment of GPRS to corroborate the statements ( see fingerpirnt case of R .v. Peter Kenneth Smith http//www.bailii.org/cgi-bin/markup.cgi?doc=/ew/cases/EWCA/Crim/2011/1296.html ) made in the ACPO "protected" document, that is the one you mentioned at the outset?

Update

GPRS CSA - http//cellsiteanalysis.blogspot.co.uk/2013/11/gprs-csa.html

Thanks for the reply. I am sorry I couldn't reply sooner.

How many of those analysts that you mention say this can actually demonstrate a deep-level independent, objective and impartial assessment of GPRS to corroborate the statements ( see fingerpirnt case of R .v. Peter Kenneth Smith http//www.bailii.org/cgi-bin/markup.cgi?doc=/ew/cases/EWCA/Crim/2011/1296.html ) made in the ACPO "protected" document, that is the one you mentioned at the outset?

I have assumed that various analysts have made their own independent analysis regarding GPRS cell usage, however I cannot find any proof of definite assessment as you say.

The question here is "Is the date/time shown with cell usage for GPRS usage accurate? Did that cell usage occurred at that exact time?" Is this a simple Yes or No question or does the answer require much more elaborate analysis as indicated in your blog?

This is what I found - For e.g. If there is a GPRS usage at 123015hrs utilising a cell-12345 and later there is also another GPRS usage at 131510hrs again utilising the same cell-12345, then the second record cannot be relied upon because the cell id indicated in the second record is a duplicate from the 1st record. Please note that there is other evidence (CCTV, ANPR etc.) which shows that the phone have moved away from the area served by the cell-12345.

This would indeed raise another question- "Is the date/time shown for each NEW cell usage for GPRS data accurate and the date/time associated with subsequent duplicate GPRS cell usages cannot be relied upon?."

I hope the above makes sense…

Hi subujoseph, thanks for coming back and it is appreciated the time and effort you have put in to this discussion. The observations I am making are not to prove anyone or any document right or wrong. What I didn't want to make anyone feel I am lecturing them about information s/he may already know, hence why I asked my question about deep-level assessment earlier.

I am putting the GPRS info first in the post and will respond in a second post to the points you have raised.

GSM GPRS

As a brief, general overview of GPRS we know that it (GPRS) is a service layer provisioned and deployed on top of GSM basic services and transmitted through GSM radio signals (cell coverage) between device and base station. GPRS is taken to be automatcally embedded with GSM unless proven otherwise. In the context of 'proven otherwise' I suggest the following observations, but this should have been checked and clarified before starting CSA or interpreting billing/call detail records/charging data records etc

1) The subscription does not allow a service between (U)SIM and Network (HLR etc).
2) The mobile device has not the capability - ~Classes A, B, C.
3) The BTS/Network is not provisioned for GPRS communications from/to the mobile station (SIM/Terminal).
4) A combination of the above.

~Classes A, B, C
(i) A Class A MS allows simultaneous CS and PS connections.
(ii) Class B MS provides automatic choice of CS or PS connection, but only one at a time.
(iii) Class C MS only supports PS connection.

The GPRS embedded device is characterised by three known 'states' for GPRS Mobility Management (GMM) for location purposes

(i) idle state - no location information
(ii) standby state - location information consisting of routing area identity
(iii) ready state - location information consisting of ^cell identity and ^routing area identity

^The cell is the area covered from a single BTS (omnidirectional or sectored). The location structure of the routing area overlays all the cell location structure and one routing area might cover many cells. The identity of the routing area can adopted to define the radii of a single cell or can define the combined cells for which it has been adopted. Thus consideration of coverage area/s in metres or kilometres is a prerequisite to understand the geographical area associated with the routing area identity.

==============
|||Routing Area ID|||
==============
_______
|||Cell |||

or

============================
||||||||||||||||Routing Area ID|||||||||||||||
============================
_______ _______ _______ _______
|||Cell ||| |||Cell ||| |||Cell ||| |||Cell |||

GPRS communications are characterised by Session Management (SM) and here again three states are identified

(i) detached state - not visible to network and no GPRS communications possible
(ii) *attached state - PDP context deactivated, routing area known
(iii) attached state - **PDP context activated, routing area and cell known

*GPRS attach procedure is initiated by the mobile device.
**GPRS PDP Context prepares the device for data communications. PDP context when activated is used for packet communication sessions within the SGSN.

Some additional points to consider
Subject to the GMM state of the mobile device it should report changes in its
location to the network.

(i) standby state - the mobile device should perform Routing Area Update (RAU) to inform the network when crossing the border into a new routing area. The mobile device after a period of being idle may perform periodic routing area updates configured by the operator which might occur 30-mins, 1-hr, 2-hrs etc depending upon requirement to keep signalling to a minimum.

(ii) ready state - the mobile device should inform the network, where necessary, of any cell reselection (CR) occurance relevant to every cell change. It should correspond that where a new routing area has been notified to the network it implies a cell reselection has occurred, too.

This brief but general overview above refers solely to GSM/GPRS and therefore eGPRS/EDGE, GPRS in UMTS etc has not been discussed. The purpose of highlighting the above is as mentioned on many occasions, and that is, it is alot easier to start with GSM (& GPRS etc) and to then comprehend how things have changed with later technology operated in the mobile network. Moreover, if you think I have mis-stated something with GSM/GPRS then, by all means put me right, but better it is dealt with now as opposed to trying to discuss it with later technology.