re: Fixed/Mobile Handover Vexes CarriersThe following snippet continues the misinformation being promtoed about UMA and IMS. "UMA provides two-way handover, but Donegan notes that neither that technology, nor any of the other current dual-mode solutions being used, are wholly compliant with the IMS standard."
Strange how an analyst, promoting a report that purports to be accurate, would make a claim the UMA is not wholly compliant with the IMS standard. UMA is not in conflict and is not "not wholly compliant" with the standard. Instead it is a definition of services and an access protocol that fits within the IMS scheme, just not as a native IMS application (such as is not SIP based). But it is not "not compliant."
re: Fixed/Mobile Handover Vexes CarriersQuestion: Where does Stoke (www.stoke.com) fit in here with their draft: draft-yafan-fmc-arch-00.txt
Comment: As I understand, BT Enterprise Fusion is not UMA, it is SIP-based. There is no hitless handover between licensed and unlicensed domains.
Response: While IMS apps (and SIP signaling) can run GÇÿtransparentlyGÇÖ over UMA as a data service, I donGÇÖt think it was unfair to say that, in terms of native support, UMA is not compliant. For example, I donGÇÖt think it was the goal of IMS developers to have 2 (or 3) distinct signaling-planes on the handset
Are you sure BT Fusion is not UMA based? Do you have any prove?
All the information passed by BT and MT through the media is that BT Fusion is UMA based.
As far as I know, IMS/SIP based solution is too costly for the provider to operate as it waste a lot mobile core network resource to anchor the call. I also don't see there is any prospect of its deployment in near or even middle term future. I really doubt your claim.
re: Fixed/Mobile Handover Vexes CarriersAs a user of the BT Fusion service for over a year here are the details of the system and the main F2M issue that they still have not solved in nearly a year.
BT Fusion works using the BT Hub and BT Home Hub this connects to the Handset using Bluetooth. The Hub converts the call to SIP (VoIP) at the CPE (Home Hub)and transports the calls over the Broadband connection.
The big issue that has existed from day one is the handling of Text Messages. When the phone is connected to the Hub certain types of Text Messages such as those from my Bank which I receive each day are not received until I spend some time away from the Hub connected to the Mobile Network which is provided to BT by Vodafone.
I recently spoke to BT Fusion again about this issue and when I mentioned it the Engineer's first words were "Tell me About It" (probably similar to SNAFU for those not used to English colloquialisms).
As I understand it the problem is in the IN platform provided by Motorola who have yet to find a workable fix. (Possibly because they need to modify the handset in some way.
I am waiting until December when my contract runs out to see if BT have a new phone that solves this issue although the only phones on offer are Motorola ones.
re: Fixed/Mobile Handover Vexes CarriersJust a comment and response:
BT Fusion is UMA based, but the enterprise version is going to be based on something "pre-standard" (proprietary?) from Alcatel. It's not clear what handsets will support Alcatel's service. That's always the challenge with non-standard, there is nothing for all the handset vendors to build too.
As for the comment that "UMA is not compliant [with SIP]", that's like saying UMTS is not compliant with SIP. It doesn't make sense.
UMA is a access/transport technology like UMTS, GSM, DSL, WiFi, WiMAX and so on. SIP runs over UMA just like it runs over any other access technology.
As for "signaling plane on the handset", today call control is managed through GSM. Tomorrow it may be managed through SIP.
But UMA is not a call control protocol, UMA simply carries the call control over IP to the mobile core to be handled.
re: Fixed/Mobile Handover Vexes Carriers[email protected] tells us: But UMA is not a call control protocol, UMA simply carries the call control over IP to the mobile core to be handled.
Jeez. Is this frighteningly off.
UMA is a method for tunneling legacy 2G signaling protocols and media over an IP network. If that ain't a call control protocol, nothing is. ...and it ain't SIP, neither.
As anyone in the industry knows, IMS specifies a SIP control plane. The client access protocol absolutely has to be SIP or it isn't an IMS client.
The fact that Kineto Wireless marketing has chosen to mix UMA with IMS in the same sentence doesn't change the IMS architecture. I'm fine with saying that UMA can interwork with IMS but the unworkable stupidly complex unlicensed spectrum solution moving target du jour in the IMS universe is VCC, not UMA.
Personally, I think UMA is the correct solution in the short/medium term. The 2G protocols are well-defined and the handset vendors just have to build the tunneling mechanism; not worry about signaling incompatibilities.
re: Fixed/Mobile Handover Vexes CarriersI think alchemy is right on the mark from a market/solution perspective. UMA is in the best case a short term 2/2.5G solution.
Mr. Shaw claims that UMA can be used to access IMS. While this is not a wrong technically speaking, UMA is a very inefficient way to access the IMS core through 3-4 boxes.
Moreover there is no interworking with RNC/UMTS, so all the operator can sell is more voice minutes...
BTW, this is not my personal viewpoint, I am just reporting what I see in 3GPP contributions. This inefficeincy of UMA was described in a 3GPP paper that was signed several operators and vendors including Kineto...
There was another poster, flaming the IMS methods, and claiming the backhaul overhead. Obviously this guy has not checked that most operators have VoIP trunks these days...
If you don't think there is a issue on the backhaul overhead for IMS based solution, what do you think is detering its deployment in the carriers' networks given they is now willing at least to have a tentative provisioning of the service (e.g. BT for BT Fusion and FT for Unik).
And we also know technically it is easier and cheaper to deploy IMS based solution rather than the UMA based solution.
How can you explain the UMA based solution is being chosen at the moment?
Backhaul overhead definitely is a issue for IMS based solution at the moment. That is why its only possible deployment situation I can see is in the enterprise network collocating with the IP PBX, and handover from the cellular network to the WLAN will not be provided.
On this issue (FMC), I think BT has a very clear view which I totally agree with.
re: Fixed/Mobile Handover Vexes Carriers>I'm fine with saying that UMA can interwork >with IMS but the unworkable >stupidly complex unlicensed spectrum solution >moving target du jour in the IMS universe is VCC, >not UMA.
Agreed ... if mobility and consistency of the service is important to an operator ('yes' for those coming from the MNO direction and 'no' for those coming from the VoIP world) then VCC represents an abysmally complex, late attempt to provide mobility between core networks (including one of which doesn't even exist yet in most - or i'm tempted to say all - mobile operator's networks). This is an approach that has never worked well in the past.
UMA can help solve this problem by providing mobility between access networks - the same way that a 2G call will hand over to a 3G network and vice versa, voice and data sessions will handover between 2G/3G and UMA access networks in the future (not just WiFi but WiMAX as well, perhaps).
The issue with UMA has been lack of native 3G CS/PS support ... this is now being addressed in 3GPP (see the latest GERAN work items). So a handset supporting 2G and 3G over UMA would have real mobility between networks for all services. That includes any IMS application using a SIP control plane over the GERAN/UTRAN packet bearer (which the operator has anyway to provide to run IMS apps over the macro network).
And just to round it off ... UMA is not a call control protocol ... its essentially a session management protocol (i.e. it manages the connection and communications between the handset and the UMA Controller for the UMA session, that's all). The call control protocol in a UMA network today is .. well, GSM (with no modifications).
"UMA provides two-way handover, but Donegan notes that neither that technology, nor any of the other current dual-mode solutions being used, are wholly compliant with the IMS standard."
Strange how an analyst, promoting a report that purports to be accurate, would make a claim the UMA is not wholly compliant with the IMS standard. UMA is not in conflict and is not "not wholly compliant" with the standard. Instead it is a definition of services and an access protocol that fits within the IMS scheme, just not as a native IMS application (such as is not SIP based). But it is not "not compliant."