IMS Reality Check

First of all, it’s important to understand what IMS actually is. The IP Multimedia Subsystem is a collection of standards defined within the 3rd Generation Partnership Project (3GPP) and other standards bodies, which together make up a blueprint for the next generation of telecom networks. Vendors are building core boxes to adhere to and accommodate these standards, so operators can begin to transition their networks to these new elements. Once implemented, the resulting all-IP network will be able to carry all traffic, be it voice, video or data, with the smarts to understand what kinds of packets are being transported and assign appropriate bandwidth and QoS allocations to them. This smarter network also will be access-agnostic, making it possible for subscribers to get their services via various kinds of wired and wireless networks and devices, without needing a separate account and password for each. And, most importantly, subscriber data will be centralized in one repository, available to any application at any time. The result? Services become abstracted from the core and access networks and are linked to a person rather than a device.

All of this is expected to translate into a series of benefits to the operator. Consolidating the core on IP means a significant reduction in operational expenditures overall. Creating and delivering services will become a rapid process, because the network will be architected to support any kind of traffic on any network and device — so the changes and reconfigurations needed to support new services in today’s networks simply become unnecessary. That, of course, improves time-to-market. Meanwhile, subscribers get ease of use and a simplified way to use their communications services, presumably leading to more service uptake and less churn.

All of this sounds pretty good, but operators face major challenges in transitioning their networks to IMS. It’s not going to happen overnight. Nor will it be easy or cheap to do. And there are no guarantees for new service revenue.

Click to Enlarge

A big hurdle is the fact that there is no definitive answer as to what constitutes an IMS deployment. “We need to remember exactly what IMS is: it is a reference architecture,” explains Chuck Kanupke, vice president at Personeta Inc. “As such, the specifications put forward by the 3GPP and other bodies do not provide a clear-cut definition of what an actual implementation looks like; this is the main source of confusion and hype surrounding the buzzword. Because the reference architect defines functional rather than physical network elements, it is important to somehow map exactly how these are fulfilled by individual equipment and technology vendors.”

At a minimum, a deployment would include the main control elements, he notes, and SIP application servers to run service applications. “There are also major issues surrounding the integration of the IMS network architecture with the emerging service delivery platform (SDP) paradigm that are not specifically answered by the IMS model,” Kanupke adds. There’s no guarantee whether one vendor’s IMS boxes work with another vendor’s. That makes for potential integration headaches, and will require extensive RFPs, plus testing and trialing of the network.

“We expect that IMS specifications will continue to evolve significantly as new requirements for extensions or variations continue to appear; however, it is difficult to predict how the various ongoing, parallel efforts at 3GPP, ETSI TISPAN and PacketCable will be harmonized in this timeframe,” says Indu Kodukula, vice president of service delivery platform at Oracle Corp. “The availability of full-fledged IMS implementations will be very closely tied to the finalization of the relevant specifications (many of which are still evolving), availability of complete end-to-end systems at reasonable price points, and availability of revenue-generating data services on top of an IMS architecture.”

Even if the specifications were gelled completely, operators still would face the problem of justifying capital expenditures to transition to IMS, when existing networks have yet to be amortized. “The IMS architecture provides operators with a re-usable, standards-based framework — enabling them to more quickly and cost effectively develop and deliver new, multimedia services to their subscribers,” says Vince Lesch, assistant vice president of product marketing at Tekelec Network Signaling Group. “But in the beginning of the transition, there will be very few IMS subscribers generating very little revenue. In contrast, there will be a lot of revenue coming from existing subscribers. Many operators view a gradual migration to IMS at the signaling layer as the most realistic, technically feasible and cost-effective approach.” That means carriers will be managing a hybrid SS7/SIP signaling network for years to come; and because they have a need to preserve existing investments in TDM infrastructure and services, they need to seek answers as to how they effectively will mix best-of-breed applications and features. “A carefully planned transition strategy — both from a network and business perspective — will be a key competitive advantage for operators migrating into the IMS space and will likely be the difference between making the IMS business case work or not,” Lesch adds.

While the network side of things presents its set of challenges, the OSS/BSS infrastructure to support the transition to IMS represents what could be the most difficult barrier of all. Service providers must retool their organizations to remove traditional boundaries between IT, network and product management/marketing functions. “The challenge for OSS to provision and service-assure the offerings will be tremendous, because of the sheer scale,” says Avner Amran, vice president of marketing and business development at TTI Telecom, which provides the Proactive Service Assurance for IMS networks. “You have one large, consolidated network with potentially hundreds of services deployed on it.”

The challenge also applies to billing, order management and customer relationship management systems. “OSS/BSS is a huge problem,” says Sita Lowman, Nortel Networks Ltd.’s IMS business leader. “There is a tremendous operational complexity involved in this that hasn’t been fully understood.”

As for the services being the key to recouping the build cost, that too requires a harder look. “You don’t have to have IMS before you can have advanced multimedia services,” explains Grant Lenahan, vice president of strategy for the service delivery solutions organization at Telcordia Technologies Inc. “You can apply IMS principles to what we have today. But the question is, as always, what will the killer app be?”

Because subscriber preferences for advanced, multimedia services still are being guessed at, the first IMS service probably will look pretty familiar, even boring. “As fancy as we get with the underlying network, we’re still talking about voice being the primary service, in the form of VoIP,” says Lenahan. “Sure, it’s voice delivered on a more cost-effective, converged network, and it’s voice you can tie to other media and functions like parental control levels, but it’s voice just the same. The right question isn’t what the services will be that pay for IMS, it’s how to make what we have more user-friendly, quicker and easier to introduce, so the uptake will be better. And that will eventually allow more innovation on the operator side.”

“Ultimately, the IMS architecture only gives service providers part of what they need to do more effective marketing in an evolving and increasingly competitive business environment,” adds Personeta’s Kanupke. “If they do not develop the ability to better segment and target their markets, then IMS cannot do it for them.”

For all these reasons, industry forecasts range from 2009 to 2011 as the years for the first real implementations of IMS, and it likely will be a full 10 to 20 years for the industry to completely convert. Meanwhile, operators are continuing to conduct IMS trials, though some are more proof of concept than in preparation for actual deployments. “Service providers cannot jump directly to IMS,” says Rod Hodgman, vice president of marketing for Covergence. “But the first stop on the road to IMS has already been reached. Carriers worldwide have replaced, or at least augmented, their traditional TDM transport networks and Class 4 switches with more economical IP transport and softswitches. The next phase — a phase that many service providers already have begun — is delivering IP-based real-time services to end users, creating a variety of new user experiences. At the same time, service portfolios expand beyond voice telephony to include video, messaging and other presence-enabled applications — often in use all at once — over ‘toll-quality’ connections. We refer to this as ‘pre-IMS’ and we expect to see a lot of these deployments in the second half of 2007.” Underpinning those service rollouts are control and application layer functional elements that use the SIP and Diameter protocols within existing infrastructures, explicitly designed to evolve over a period of years to full IMS deployments. The last phase will introduce SIP and IMS network elements into the carrier infrastructure, which could begin in 2008 or early 2009 and to continue through 2011, consensus says.

“Our take is that people (i.e. vendors) should forget about all the acronyms and just solve the problems,” says Keith Higgins, vice president of marketing for Stoke Inc., which sells an IP service delivery platform built for converged fixed/mobile networks.

Interop Pragmatics

To address some of the interoperability concerns around IMS, industry groups are heading to the lab.

The MultiService Forum held Global MSF Interoperability 2006 last fall, a multivendor, three-continent interoperability event to validate the core capabilities of IMS, particularly the ability to support fixed-mobile convergence. British Telecom, Korea Telecom, NTT, Verizon Communications Inc. and Vodafone brought their networks together, while the University of New Hampshire Interoperability Lab and ETRI oversaw the proceedings, and sponsor Nortel and a total of 27 vendors participated.

The global network was built using the MSF Architecture Release 3 as a physical implementation specification. Vendors whose products complied with all relevant MSF implementation agreements were shown to interoperate. While it is only one way to put an IMS network together, it was at least a real-world trial of the full architecture.

This month, the IMS Forum is conducting a plugfest to scrutinize how applications interact with the network. “We will be looking at real applications under real conditions and how they will perform, including VoIP, presence and dual-mode wireless,” says Michael Khalilian, chairman and president of the IMS Forum.

Part of the benefit of such work is to lessen the systems integration requirements that operators will face. “With IMS, you now have a lot of products and servers out there,” says Sita Lowman, Nortel Networks Ltd.’s IMS business leader. “That will require a strong systems integration effort to make it work. So we’re seeing a marriage between IT and the network.”

The systems integration tax can be much greater than paying the 25 percent more for the IMS boxes, according to Grant Lenahan, vice president of strategy for the service delivery solutions organization at Telcordia Technologies Inc. “You need to put together a service delivery framework to serve IMS, IT concerns and existing wireless and wireline networks,” he says. “It’s easier if you have a road map.”

Share this article: Email, Slashdot, Digg, Del.icio.us, Yahoo!MyWeb, Windows Live Favorites, Furl
Add this article feed to: RSS, My Yahoo, Newsgator, Bloglines

Read Comments [1]