Comments on: TeliaSonera Taps Kineto for WiFi-Cellular http://gigaom.com/2006/09/25/kineto-teliasonera/ Trusted Insights and Conversations on the Next Wave of Technology Fri, 27 Nov 2009 00:25:55 +0000 http://wordpress.com/ hourly 1 By: Jesse Kopelman http://gigaom.com/2006/09/25/kineto-teliasonera/#comment-65675 Jesse Kopelman Thu, 28 Sep 2006 21:01:38 +0000 http://gigaom.wordpress.com/2006/09/25/kineto-teliasonera/#comment-65675 <p>Ulf and Troy, thanks for the info. I guess I always assumed Kineto had built SIP into UMA since that is a perfectly valid protocol for what they were trying to do. That said, I'm not surprised they went a more proprietary route. While WiMax, IMS, and other efforts may change things, Carrier Grade has usually meant proprietary (in the sense that while standards are employed they are done in such a way as to limit cross functionality and tie one to a specific platform).</p> Ulf and Troy, thanks for the info. I guess I always assumed Kineto had built SIP into UMA since that is a perfectly valid protocol for what they were trying to do. That said, I’m not surprised they went a more proprietary route. While WiMax, IMS, and other efforts may change things, Carrier Grade has usually meant proprietary (in the sense that while standards are employed they are done in such a way as to limit cross functionality and tie one to a specific platform).

]]> By: Troy Pummill http://gigaom.com/2006/09/25/kineto-teliasonera/#comment-65674 Troy Pummill Thu, 28 Sep 2006 16:59:56 +0000 http://gigaom.wordpress.com/2006/09/25/kineto-teliasonera/#comment-65674 <p>Mr. Nilsson is correct; UMA does not employ SIP. As Ulf, I work closely within the UMA technical world and am the principle author for several patents for UMA applications and technology.</p> <p>UMA, in essence, is GSM over UMA over IP. The design goal for UMA is to provide a transport mechanism between the MS and the core without changes to either. By tunneling the GSM protocol over UMA, neither the GSM handset nor the cellular core require any changes. The UNC binds the IP infrastructure to the cellular infrastructure providing the necessary bits to seamlessly glue to the two together (it provides a cellular instantiation for the IP-based UMA phone).</p> <p>The concept of carrying IMS over UMA is workable and has been demonstrated. It may, or may not, be a desirable solution. Certainly once the UMA infrastructure is installed, using it as an Access Technology (which it excels at) is a compelling option.</p> <p>IMS, IMHO, has bigger issues than just how it will be accessed. SIP, on which IMS is built, is still undergoing massive changes - interoperability problems are still the norm. Take a shaky foundation and build an entirely new, massive infrastructure on top of it and you're asking for trouble.</p> <p>In the networking world, simplicity is king. Why do we have SNMP instead of CMOT? Or TCP/IP instead of ISO TP? UMA is simple, cost effective, and has high interoperability. The number of applications and services that UMA can provide has yet to be revealed. Keep your ears open!</p> Mr. Nilsson is correct; UMA does not employ SIP. As Ulf, I work closely within the UMA technical world and am the principle author for several patents for UMA applications and technology.

UMA, in essence, is GSM over UMA over IP. The design goal for UMA is to provide a transport mechanism between the MS and the core without changes to either. By tunneling the GSM protocol over UMA, neither the GSM handset nor the cellular core require any changes. The UNC binds the IP infrastructure to the cellular infrastructure providing the necessary bits to seamlessly glue to the two together (it provides a cellular instantiation for the IP-based UMA phone).

The concept of carrying IMS over UMA is workable and has been demonstrated. It may, or may not, be a desirable solution. Certainly once the UMA infrastructure is installed, using it as an Access Technology (which it excels at) is a compelling option.

IMS, IMHO, has bigger issues than just how it will be accessed. SIP, on which IMS is built, is still undergoing massive changes – interoperability problems are still the norm. Take a shaky foundation and build an entirely new, massive infrastructure on top of it and you’re asking for trouble.

In the networking world, simplicity is king. Why do we have SNMP instead of CMOT? Or TCP/IP instead of ISO TP? UMA is simple, cost effective, and has high interoperability. The number of applications and services that UMA can provide has yet to be revealed. Keep your ears open!

]]> By: Ulf Nilsson http://gigaom.com/2006/09/25/kineto-teliasonera/#comment-65673 Ulf Nilsson Thu, 28 Sep 2006 06:39:19 +0000 http://gigaom.wordpress.com/2006/09/25/kineto-teliasonera/#comment-65673 <p>Just for the record, I worked quite a bit on the technical side of UMA within TeliaSonera. UMA is proprietary (although now standardized by 3GPP) in the sense that it has nothing to do with SIP; the signaling between the handset and the network is an IP-fied version of the signaling done in the cellular macro network keeping the core network signaling/messages intact. </p> <p>The path towards IMS does therefore not necessarily go over UMA-enabled networks. One possible way, that is being promoted by Kineto, is to run IMS on top of UMA, i.e. go though the mobile PS domain before hitting the IMS, which seems to waste a lot of bandwidth on protocol overhead. The big advantage of UMA is instead rolling out the basic mobile service of today on mainly WLANs, giving the user the same cellular-network services (i.e. no new fancy stuff). On the other hand, the services, supplementary or SMS and similar, will work from day one - something that is not obvious in early IMS deployments.</p> Just for the record, I worked quite a bit on the technical side of UMA within TeliaSonera. UMA is proprietary (although now standardized by 3GPP) in the sense that it has nothing to do with SIP; the signaling between the handset and the network is an IP-fied version of the signaling done in the cellular macro network keeping the core network signaling/messages intact.

The path towards IMS does therefore not necessarily go over UMA-enabled networks. One possible way, that is being promoted by Kineto, is to run IMS on top of UMA, i.e. go though the mobile PS domain before hitting the IMS, which seems to waste a lot of bandwidth on protocol overhead. The big advantage of UMA is instead rolling out the basic mobile service of today on mainly WLANs, giving the user the same cellular-network services (i.e. no new fancy stuff). On the other hand, the services, supplementary or SMS and similar, will work from day one – something that is not obvious in early IMS deployments.

]]> By: Jesse Kopelman http://gigaom.com/2006/09/25/kineto-teliasonera/#comment-65672 Jesse Kopelman Wed, 27 Sep 2006 22:21:13 +0000 http://gigaom.wordpress.com/2006/09/25/kineto-teliasonera/#comment-65672 <p>Jacomo, UMA is currently GSM only so when you talk about US carriers waiting for IMS you are really just talking about Cingular. Cingular is pretty damn conservative. Last week their Network Honcho Ed Reynolds shocked everyone at a backhaul conference with how he had no desire to move away from TDM backhaul to Ethernet/MPLS (meanwhile Sprint is trying to do just that as fast as they can). As to an upgrade path from UMA to IMS: The whole point of UMA is that it uses a gateway to translate SIP to/from the phone into 3GPP signaling at the switch. With IMS you have SIP inside the switch, so no need for gateways. The real question is whether the SIP implementation on the UMA handset is compatible with IMS (some have claimed it is not really SIP but a similar proprietary session protocol, but I've not heard an answer from Kineto)?</p> Jacomo, UMA is currently GSM only so when you talk about US carriers waiting for IMS you are really just talking about Cingular. Cingular is pretty damn conservative. Last week their Network Honcho Ed Reynolds shocked everyone at a backhaul conference with how he had no desire to move away from TDM backhaul to Ethernet/MPLS (meanwhile Sprint is trying to do just that as fast as they can). As to an upgrade path from UMA to IMS: The whole point of UMA is that it uses a gateway to translate SIP to/from the phone into 3GPP signaling at the switch. With IMS you have SIP inside the switch, so no need for gateways. The real question is whether the SIP implementation on the UMA handset is compatible with IMS (some have claimed it is not really SIP but a similar proprietary session protocol, but I’ve not heard an answer from Kineto)?

]]> By: Jacomo http://gigaom.com/2006/09/25/kineto-teliasonera/#comment-65671 Jacomo Tue, 26 Sep 2006 17:18:06 +0000 http://gigaom.wordpress.com/2006/09/25/kineto-teliasonera/#comment-65671 <p>Will be interesting to see what the Big USA based CellCOs will do when T-Mobile, now rich with AWS bandwidth, begins deploying the UMA based COnverged service offering and new Dual Mode handsets locally. CIngular has released some Dual Mode Smartphoens but is not providing any convergence yet. I do not expect much short of more long term plans to wait until IMS is available.</p> <p>MY question is, why are the Eurpean Carriers not waiting for IMS and going ahead with UMA? Is this an issue with SIP based VoiceIP systems for USA?? Doesn't Kineto provide a SIP based transition plan too?</p> <p>Jacomo</p> Will be interesting to see what the Big USA based CellCOs will do when T-Mobile, now rich with AWS bandwidth, begins deploying the UMA based COnverged service offering and new Dual Mode handsets locally. CIngular has released some Dual Mode Smartphoens but is not providing any convergence yet. I do not expect much short of more long term plans to wait until IMS is available.

MY question is, why are the Eurpean Carriers not waiting for IMS and going ahead with UMA? Is this an issue with SIP based VoiceIP systems for USA?? Doesn’t Kineto provide a SIP based transition plan too?

Jacomo

]]> By: AP http://gigaom.com/2006/09/25/kineto-teliasonera/#comment-65670 AP Tue, 26 Sep 2006 07:03:49 +0000 http://gigaom.wordpress.com/2006/09/25/kineto-teliasonera/#comment-65670 <p>Looks like WSJ today just reporting some exciting news from T-Mobile with more related offerings of plans and devices(hopefully national) in the month of Oct. http://online.wsj.com/article/SB115923297434173806.html?mod=rss<em>whats</em>news_technology</p> <p>Definately curious, like Katie, for first hand user reports.</p> Looks like WSJ today just reporting some exciting news from T-Mobile with more related offerings of plans and devices(hopefully national) in the month of Oct. http://online.wsj.com/article/SB115923297434173806.html?mod=rsswhatsnews_technology

Definately curious, like Katie, for first hand user reports.

]]>