Just summarizing my panel session speech on day 3 of SofNet, “A Pragmatic Approach to Developing IMS-Services.”  I addressed two questions I’m often asked.

“Is IMS dead?”  Both Sprint and Verizon are deploying their flavors of IMS for the simple reason they decided voice over EVDO (the CDMA version of 3G) will be supported by VoIP, so they needed a standard IP session control layer, to which IMS provides a solution.  I discussed this previously in the “The Divergence of CDMA and UMTS Operators on IMS” weblog entry.  By ‘flavor,’ I mean they are using some of the specifications within IMS as a guide, its not a ‘soup-to-nuts’ implementation of the current IMS specifications.  Now NTT and SKT have taken an aggressive approach to deploying IMS, with a number of services including PTT (Push To Talk) and virtual PBX, but they traditionally adopt technologies early to give their national suppliers an international competitive edge with early reference customers and accelerating their national suppliers’ position on the learning-curve.

For UMTS operators voice is supported as a circuit over the RAN (Radio Access Network), so there is no need for an IP session control layer, yet.  Hence most UMTS operators have not deployed IMS.  On the fixed side, most operators have implemented softswitch solutions, and in some cases there are scalability issues, there they are evaluating whether IMS provides a more scalable solution, but the key word is evaluating.  So IMS has not failed, rather its core capability of IP session control is seeing pragmatic deployment, where it makes business sense.  Now, standards bodies as always tend to go beyond their initial scope, so the standards engineers can maintain the frequent flyer status, and IMS has moved into service management aspects.

This leads to the second question I’m often asked, “Are IMS and SDP in competition?”  Immediately we have the problem of definition, what is an SDP?  Paraphrasing what Humpty Dumpty said to Alice, “It means whatever I chose it to mean.”  In practice there are three main elements, BOSS (Business and Operational Support Systems), Content Delivery and the Telco API (checkout this entry for more info “The Telco API: Potential to raise ARPU by up to 36%”).  Here we have lots of deployment examples such as Sprint’s business mobility framework, AT&T’s U-Verse, Airtel’s managed SDP, Vodafone Live! etc.  So the SDP is happening, and is more successful than IMS.  Its focus is service management, not IP session control.  So at present they are not competing, and if the focus of IMS remains IP session control then they will not compete.

But back to the topic of the panel, IMS-Services.  If someone starts talking about IMS or SDP-Services, they have an agenda, they’re trying to sell or promote a network technology, and making a fundamental error in looking at the problem from the wrong-end.  Services are supported by capabilities, and those capabilities are implemented with network technologies.  Services come become technology, as technology enables implementation.  Of course user experience comes first, and then services fulfill the experience, but that’s covered in the “A Critical Gap that Means most Service Platform (IMS/SDP) Business Cases Keep Failing” entry.

Overall the conference was worth attending, though there was a significant gap around specific enabled services.  To claim the conference is focused on Software and Networks will require much greater participation from the other members of the value chain, including media, web 2.0 companies, and internet application developers.