IMS: Revival or Last Stand?

There are few, if any, technologies that have managed to polarize the industry as much as IP Multimedia Subsystem (IMS) has and still survive.

On the one hand, detractors paint IMS as having missed the window of opportunity and lacking a business case. On the other, supporters highlight Rich Communications Suite (RCS) commercial launch commitments and Voice-over-LTE (VoLTE) announcements as proof that IMS is once again trending positive.

But are these developments simply a last stand for IMS, or the beginning of a revival? Even after seemingly endless years of waiting, it will realistically take at least another 18 months to determine if these initiatives are game-changers. (See IMS: As Seen on TV.)

So in the meantime, let's review how we got to this stage.

One of the arguments often cited in the demise of IMS is its complexity. And without question, IMS is a complex beast. I can only imagine the tone of discussions between network planners and executives using a generic IMS architecture diagram -- with the outcome reached that it was necessary to implement all these nodes and interfaces to be IP-ready and hope for services down the road.

In addition, at a time when 4G standards advocate combining nodes in the radio access network to decrease latency versus further distribution, if the industry defined an IMS architecture today, would it be so distributed? The dilemma is that if you don't, you basically have a softswitch, which is not optimally suited for converged carriers that require a single core network.

Another standards-related consideration is interworking. While IMS was predicated on defining open interfaces to enable vendor interworking and some limited vendor plug and play has taken place, for business and technical reasons IMS is still often delivered by a single vendor, which limits services interoperability, and results in the very "lock in" it was conceived to avoid.

Although vendor approaches such as IMS on a blade may simplify implementation and address smaller carrier requirements, they don't address this concern. Moreover, while the industry has taken steps to address these interoperability concerns through approaches such as the RCS initiative and activities of the IMS Forum , progress has been slow.

There are also service strategies to consider: While IMS was touted as providing a framework for IP services, in reality, the early focus was on what I define as "small i" IP services, and not compelling. I can recall walking around GSM World Congress in 2006 thinking, "Is everyone really going to use push-to-talk?"

While trying to cultivate and sell a primarily voice-centric application such as push-to-talk was clearly a mistake, and only served to embellish IMS's service limitations -- what other choices were there? At the time -- given the IP services model had yet to be shaped by industry trends -- not many.

Therefore, isn't one of the key reasons IMS faltered out of the gate simply because it was too early to market?

For example, although UMTS was starting to roll out in 2006, GSM still had a great deal of growth ahead of it, particularly in emerging markets. And factoring in that WAP 2.0 Web browsing services -- while not perfect -- could be supported on these networks, the case for justifying an IMS IP core for TDM networks was a non-starter. If a carrier did require IP connectivity, a softswitch was more desirable because it was still a "switch" and supported a similar feature set.

Moreover, there are the terminals to consider. Even if there was a much greater penetration of UMTS networks, the performance capability of vintage feature phone and early 3G terminals would have detracted from the overall IMS end-user experience.

While being too early to market is one plausible explanation, the telecom landscape is strewn with visionary ideas that were passed over once and never afforded a second chance. So what's changed for IMS this time around? Based on the above, I would argue almost everything -- especially for mobile operators that will drive future IMS growth.

Let's start with a greater penetration of IP in access and backhaul networks of mobile operators. There is little doubt that UMTS/HSPA over the past few years has become the dominant technology, and more importantly provided operators with the initial tools and connectivity to start delivering IP services.

Similarly, the terminal design innovation, performance, price points and ultimately adoption curve of smartphones over the same period are other factors that enhanced and validated the IP service experience, instead of detracting from it.

As a result, somewhere in the past few years, there has been a fundamental shift in how IP technology is viewed. When talking to mobile and fixed operators as recently as mid 2008, there were still a significant number that defined IP as disruptive technology.

Since then, with the unprecedented adoption of social networking services, IP technology has achieved mainstream status and will only gain momentum as LTE is rolled out. And while telco-driven open application programming interface (API) initiatives designed to compete with over-the-top (OTT) applications don't specifically require IMS, there could be increased value in leveraging this technology to scale and integrate applications as well.

Given that IMS was designed with the fundamental assumption that IP, not TDM, would be the dominant underlying technology, and that we are finally (despite some missteps) reaching that point on the horizon, it's now actually conceptually more attractive.

The outcome of recent initiatives such as RCS, VoLTE and even IMS pricing models require further study and scrutiny before a definitive decision can be reached. But for IMS, the old adage "third time's the charm" might actually ring true.

— Jim Hodges, Senior Analyst, Heavy Reading