From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============6969717668609323342==" MIME-Version: 1.0 From: Denis Kenzior Subject: Re: [PATCH] atmodem: CEREG support for LTE network status reporting in AT modem Date: Thu, 24 Feb 2011 10:12:37 -0600 Message-ID: <4D668375.7040604@gmail.com> In-Reply-To: <4D663426.7020706@nokia.com> List-Id: To: ofono@ofono.org --===============6969717668609323342== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Hi Arun, On 02/24/2011 04:34 AM, Arun Ravindran wrote: > Hi Denis, >>>> -----Original Message----- >>>> From: ofono-bounces(a)ofono.org [mailto:ofono-bounces(a)ofono.org] On >>>> Behalf Of Denis Kenzior >>>> Sent: Wednesday, February 23, 2011 5:50 PM >>>> To: ofono(a)ofono.org >>>> Cc: tomasz.gregorek(a)gmail.com >>>> Subject: Re: [PATCH] atmodem: CEREG support for LTE network status >>>> reporting in AT modem >>>>> After discussing internally with Fred Joly I would like to come back >>>> on the 4 properties : >>>>> - Suspended is applicable in LTE in case of CS Voice Call using CSFB >>>> My understanding was that CS Fallback physically switches technologies. >>>> It is not 'suspending' LTE like it does with GPRS. Am I wrong on >>>> this >>>> one? >>> I would say yes. :) >>> In case you are obliged to fallback from 4G to 2G without support of >>> DTM, exaclty like in GPRS. >> From what I understood you physically lose the EPS network registration >> and fall back to 3G / 2G. How is this the same as the suspended >> property where your attached status does not change? >>> > For CSFB there are two settings. First one is that the UE should be > voice centric. > A voice centric UE will always try to make voice service available > either using IMS or by CSFB. > And if it cannot avail voice service using the above mechanisms, then it > should leave E-UTRA and re connect to GERAN/UTRA. A data centric UE in > turn will not leave E-UTRA if it cannot make a voice service available. > (23.221, section 7.2a) > = > The second is that the voice call mode (domain), this can be set to CS, > CS preferred, PS or PS preferred. A UE with CS or CS preferred will > always try to make voice service through CS (CSFB or by re connecting to > GERAN/UTRA). 24.167 and also 23.221 describes these setting. I think the > AT commands for these in 27.007 are AT+CEMODE and +VCMOD > = > A UE indicates its CSFB capability in attach and location/routing area > updates. The network in turn also indicates whether these are supported > in network. > 23.272 section 6.2 details a CSFB for MO call, where PS handover is > supported. 6.3 where PS handover is not supported and 6.5 where how to > return to E-UTRA when PS is suspended. > = So everyone keeps talking of suspensions. However, I still see no CGEV that supports this event. It still looks like a 2G holdover that is being implemented by vendor specific extensions. Or is this just an oversight on 3GPP's part and suspensions can happen in LTE as well? > My understanding is that the context behaves the same way as in GPRS, > where it can be suspended, deactivated or handed over while doing CSFB > and the attach status will not change because when you are doing CSFB > you have done a combined attach (EPS/IMSI). > = Attach status fair enough. However, does the reported technology change as well? There are important implications here, e.g. being able to deactivate the last context, etc. Overall I think we need detailed traces for all such scenarios and consider all possibilities. Currently I'm not convinced sharing LTE and GPRS logic is the right way to go. >>>> Perhaps, but again we have no control of the attach procedure with LTE. >>>> So the implementation has to rely on vendor specific radio settings. >>>> Same arguments about shoehorning logic apply here. >>> I'm a bit puzzled by this statement. It is not specific to LTE, but >>> related to +CFUN proprietary implementation against +COPS or +CGATT = >>> usage, isn't it? >>> >> The current GPRS logic is controlled with the equivalent of CGATT. This >> is a well-defined command by 3GPP, but it is not applicable to EPS >> domain. It is applicable only to PS domain, correct? > 27.007 release 10, section 10, lists all commands applicable to EPS and > CGATT is one of them. > = My fault, I mis-interpreted that section. You're of course correct that CGATT is indeed applicable. Regards, -Denis --===============6969717668609323342==--