By Brough Turner, SVP & CTO
Brough Turner, recently interviewed by Rafael Junquera, editor at Tele-Semana , provides his outlook on IMS—where we are and its impact on fixed and mobile telecommunications. The following Q&A will be featured in Tele-Semana ’s August report on IMS ( www.tele-semana.com ).
1. How would you describe IMS in 300 words maximum?
As with any generational change, the term IMS (IP Multimedia Subsystem) has acquired a spectrum of meanings. The original definition of IMS comes from the 3rd Generation Partnership Project (3GPP) — the organization developing standards for the 3G evolution of GSM mobile networks. IMS specifications first appeared in the 3GPP Release 5 (R5) as the way to support multimedia communications over IP. In this context, “multimedia” includes voice, so from the beginning, IMS has been positioned to replace traditional circuit-switched voice telephony as well as support new multimedia services.
The key technical feature of IMS is the use of SIP (Session Initiation Protocol) to set up individual IP sessions under control of a “Call Session Control Function” (CSCF), otherwise known as a softswitch. This allows IMS to guarantee IP quality of service, provide security and support billing on a per session basis.
Today the term IMS is used in many ways. Sometimes it’s used when an operator has added any kind of SIP infrastructure to their network and sometimes it refers to a network that has actually deployed equipment conforming to 3GPP R5 or Release 6 (R6) specifications.
Of course with any major technology change, the hype cycle gets out in front of actual deployments. Further, the IMS specifications themselves have been evolving as 3GPP R6 filled in gaps that made R5 difficult to actually deploy. Then, IMS attracted the attention of other standards groups including TISPAN which focuses on next generation fixed networks and CableLabs which focuses on cable operators and their telecom interests. 3GPP Release 7 (R7) represents the first IMS designed to include all kinds of mobile and fixed networks.
2. Why is IMS important for mobile operators?
Pre-IMS mobile networks are voice-centric—even 3G video telephony (3G-324M) operates over circuit-switched data paths. But mobile radio technology is beginning to approach broadband data rates. If the operators don’t move to support IP-based services, others will provide such services, as best they can, using bulk Internet access.
3. Why is IMS important for fixed operators?
Fixed operators are being squeezed by mobile and by VoIP arbitrage. At a minimum, they need to support fixed-mobile convergence, but ideally they need a way to deliver new services. IMS has the potential to support new services, with quality of service guarantees that voice over the public Internet cannot offer.
4. Is IMS more critical for carriers that have both fixed and mobile arms?
IMS is a good way to support fixed-mobile convergence (FMC). There are other approaches (like UMA), but they are service specific whereas IMS is a general platform for multiple services.
5. What are the major challenges faced by IMS?
First is complexity. There are many, many IMS specifications, they come in multiple versions which continue to evolve, and interoperability testing hasn’t kept up with product introduction and deployment.
Second is services. IMS is a platform that can support new services, but the operator doesn’t make money until services are launched and succeed. Today, there are a relatively small number of new services available with IMS and, compared with the Internet, a small developer ecosystem.
6. Is there an alternative to IMS?
The major competition is services implemented directly over IP and/or over the open Internet. Such services can’t rely on IMS’s quality of service, but the pool of developers is vast, and IP quality of service guarantees are not always needed.
7. How would you describe the current status of IMS?
The majority of today’s “IMS” deployments are a mixture of SIP elements (IMS-compliant or otherwise) and legacy networks. Since IMS for fixed networks only begins with R7, today’s fixed “IMS” networks are typically softswitch controlled voice circuits plus some kind of SIP infrastructure for new applications like instant messaging (IM) and presence or fixed-mobile convergence.
To the extent true IMS networks are deployed—they are being used for new applications like IM, FMC and video sharing, but not yet for basic voice telephony.
8. Which operators around the globe are starting to implement IMS? Are the big players the first ones to implement IMS?
IMS capabilities are showing up in major networks around the world and in green fields networks, for example in the Middle East.
9. What is the state of standardization?
The R7 specifications have been functionally frozen but are still not released, and some items, like IMS enhancement in support of cable telephony, have slipped into Release 8. Typically it has taken at least two years from a stable specification to equipment that can undergo interoperability testing.
10. Do you think standardization is the major obstacle faced by IMS or is the lack of compelling applications?
Compelling applications, i.e. applications that make money, are the biggest issue. Standardization and interoperability increases an operator’s options when it comes to equipment purchases, but in the end it’s only new applications that generate incremental revenues.
11. Why is Release 7 so important? Are carriers waiting for it to fully deploy IMS?
IMS R7 is the first release that integrates mobile and fixed network support. Also, with hindsight, R5 is considered incomplete, and several major changes took place between R6 and R7 (for example in the policy infrastructure) that mean any R6 deployment would require major rework to reach R7.
12. IMS implementation is an evolution, how long will it take for operators to completely migrate to it?
The IMS transition is likely to take a decade or more, just as the “Intelligent Network” (IN) did in the last generation. The complexity of IMS and IN are similar and the industry structure (vendors, operators and standard organizations) remains the same, so it’s hard to see IMS happening any more rapidly than IN.
13. How will IMS affect users? Will they have more services at lower prices?
The extent to which IMS results in many new services depends upon how open operators are to 3rd party developers. If operators continue current patterns, then IMS will do as well as IN did, i.e. a decade later IMS will have fostered the extremely wide deployment of a few new services (perhaps less than ten).
14. How critical are IMS capable mobile devices? When do you expect them to be available?
IMS capable handsets are critical for some applications, for example IMS implementations of instant messaging and presence. Other applications don’t need, or can work around the lack of, IMS capable handsets.
IMS capable mobile devices are showing up right now. Among other things, NMS is a supplier of IMS handset software to companies like Samsung. Handsets with our IMS software began arriving on the market earlier this year.
15. How will IMS impact HSPA?
HSDPA and HSUPA are radio capabilities that bring “broadband” access speeds to UMTS networks, so technically, HSPA is relatively decoupled from IMS. However, in competitive markets, the availability of HSPA increases the likelihood that one or more operators will offer flat-rate broadband Internet access, which in turn increases the likelihood that 3rd parties will implement Internet-based (non-IMS) services over mobile networks.