IMS - Its Challenges & Benefits

Buzzwords come and buzzwords go, and that’s certainly the case with IMS. A year ago, the IMS Forum was completing its first plugfest and IMS fever showed no signs of abating. Now, IMS has taken on a lower profile. So low, in fact, that the former industry darling ended up on many "worst of 2007" lists. That may be why you don’t hear many vendors bandying the term about anymore.

But IMS is not a failed experiment — in fact, far from it. The march toward consolidating traffic on a converged IP architecture that uses SIP to enable rapid creation and delivery of enhanced services is happening. Just not as quickly as we were all told that it would. You could characterize it as a technological movement that has gone from hot to simply percolating — it’s getting ready to be served.

"IP is a new technology for operators as it will, inevitably, replace their existing network," explains Victor Donselaar, president of Movial Applications, which creates mobile and IMS-based software clients. "Basically, at some point, IMS or something that is a standardized core and application network based on IP (today IMS is the only real contender), will replace the existing core network. This is a radical step for any operator that has for years and years made investments in circuit-switched telecom networks. Considering the billion dollars of investments in circuit-switched and non-IP technologies, I think it’s a wonder that more than 50 IMS cores have been sold already in 2007."

The cooling steam of the formerly hyped IMS conceals the fact that in 2008 vendors and service providers alike are forging ahead, conducting trials, demonstrations and rollouts. Over the course of the last year, Ericsson alone claimed 39 commercial IMS contracts as of September 2007, with a total number of IMS contracts worldwide from all vendors topping 50. "That isn’t bad for a new technology," Donselaar says. "If you calculate the pre-IMS deployments, this figure will be a lot higher." Veraz Networks reports it is seeing new service providers deploying full services over IMS, and Veraz has two customers in this category. Meanwhile, ZTE USA Inc. is deployed with China Mobile Ltd. In the States, AT&T Inc. has made IMS transition a priority.

Vendors also say that more IMS-based RFPs are appearing in the market. Majid Foodeei, director of strategic and technical marketing at IMS-on-a-chip maker Centillium Communications Inc., says service providers are looking to IMS as a future-proofing strategy. "They want it to go toward an architecture that brings the benefits they’re after — in this case, rapid service creation and delivery," he says. "A new RFP from France Telecom is out there that’s very IMS-centered, for example. Also, we have been focused on infrastructure and CPE, and we’ve seen IMS architecture play into the way OEMs build their equipment. So we do see IMS deployments, however slow. Tier 1 service providers are after this and are safeguarding investment for future."

At the same time, the IMS Forum continues to target real-world challenges, most notably in its fourth plugfest in February, one of three events it plans for 2008. The latest plugfest is focused on OSS issues — the back office being one of the major obstacles in making IMS work properly in a live operator environment. "2008 is a huge year for public interoperability events to prove readiness," says Aaron Sipper, director of product marketing at Nextpoint Inc., which offers IMS solutions for user-to-network and network-to-network applications. "For instance, this year ETSI will be sponsoring two interoperability [events] at operator labs, the IMS Forum is holding three test events, and the MultiService Forum is holding a global multiservice interoperability event that includes IMS — it’s being held in operator labs in Asia, North America and Europe with services running across continents."

Also, the core IMS standards have been finalized and released. "Most standards development organizations are working on standardizing the IMS extensions for regional deployment issues such as emergency communication, policy control charging and lawful inception," says Weijun Lee, vice president of network architecture at ZTE USA.

IMS momentum will be affected by the fact that IP is revolutionizing the way operators have to think about their business models. "User service requirements are the most important driver for IMS," says Stu Bennington, director of global portfolio marketing for Tellabs, which is focused on enabling operator environments for traffic control, QoS and control plane. "While there are potential cost savings from converging multiple service types onto an integrated technology standard such as SIP, the more lucrative opportunity is for IMS to drive a new user experience that will ideally create a willingness for those users to pay more for new services or differentiate existing ones."

Therefore, "a major challenge to a commercially successful IMS launch is for the small, but growing, ecosystem of application partners to provide revenue-generating services," says Ken Lee, director of worldwide product marketing for the BEA WebLogic Communications Platform. BEA Systems Inc. is a provider of IMS application servers. "A key point to highlight is the fact that IMS deployments (trials, POC, commercial) are actually occurring, with some operators generating revenue," says Lee. "But another key issue is that there is still a lack of new applications that can be identified as being uniquely dependent on IMS technologies and capabilities. Some IMS operators have launched VoIP services on IMS."

Nextpoint’s Sipper says because of this, IMS needs to demonstrate that it is a multiservices architecture built to support not just SIP, but also IPTV, XML and Web 2.0 applications. "There is more thought around services than just cost reduction," he says. "What is driving the emphasis toward services is Web 2.0 and IPTV architectures. Web 2.0 and IPTV can be viewed as competitive to IMS. IMS must include provisions to integrate Web 2.0 and IPTV functions." Requirements for support for multiservices capability (such as IPTV, video on demand and gaming applications) are influencing accordingly the next revisions of IMS to support signaling and control protocols beyond SIP (and VoIP).

It’s also important to note that full IMS deployments will be slow because of market realities, plus a litany of problems that need to be solved and market development that needs to be done.

BEA'S Ken Lee

One truth is that service providers aren’t going to rip out perfectly functional network equipment, and some deployments continue to be non-IMS-based. "It was for a long time exaggerated that nothing other than IMS will get deployed from now on — and that’s just not true," says Foodeei. "Many network expansions [are] not really IMS, and that’s particularly true on the fixed side. The biggest promise of IMS is that service providers can very, very quickly try services, test market, create apps, lower costs, and to do that [there are] still a lot of challenges — many pieces that have to come together and it will take time, so why remove your existing solution in the meantime?"

Some operators have done analysis, and they can’t justify the cost for moving to IMS just yet, Sipper notes. "For a number of fixed-line only operators, there is not a clear ARPU increase or reduction in capex/opex to justify IMS," he explains. "Most of the operators that have done the analysis and have made the investment typically have fixed-line and mobile assets. The reality is that IMS is not for every operator — might work for some, and might not for others."

For these reasons and others, existing operators generally are adopting an incremental approach to transitioning to IMS, rather than any wholesale migration. Pierre Lynch, technical product manager for IMS and wireless core at test and measurement company Ixia, says the industry is concentrating first on session management, security at the edges and common databases. "Session management is the buzzword," he notes. "Everyone views it as the most important feature in a converged IP network."

Another reality to consider is there are immense complexity and lingering challenges in building out an IMS-based network. "The promises of IMS are, in fact, very attractive, but are not reachable in the very short term because of the complexity of the initial rollout, and the fact that some of the required specifications for a full-blown, all-features-ready system deployment aren’t cooked yet," says Lynch.

Then there are the management requirements to consider. "For incumbent service providers, the challenge is in determining how to generate revenue from new services over IMS versus their traditional services, and what is their service delivery model," says Amit Chawla, executive vice president of global business units for Veraz Networks. "The incumbents need a service broker, which allows a single OSS integration point, for provisioning and billing, across all application servers, whether delivered over a traditional network or an IP-based network. This allows incumbent service providers to uniformly manage all services, whether for the IP-based network or traditional network." The Veraz SCIM is one such solution.

The bottom line? IMS is unlikely to go away, despite it’s falling out of the public eye. Simply put, while convergence can be done without IMS, it can’t be done as well, "and certainly not as efficiently," says Lynch. "This is a well-known and accepted fact. A common infrastructure and mechanism for session management, subscriber database management, security and charging is simply the best way to go. Additionally, PSTN substitution using the TISPAN NGN architecture, which will be merged with the IMS architecture, is acknowledged by the industry as the best way forward."

So when will we see significant migration? "The validation of IMS in the marketplace will really come when network deployments take place on a reasonably large scale and when there is demonstrated ability to gain new revenue streams," says Bennington. "The main change is consortiums of vendors and service providers are advocating modified standards to address some of the real-world deployment considerations that they see."

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