IP multimedia subsystem, or IMS, architecture seems like much more of a meaningless mouthful than unlicensed mobile access, but UMA seems to be the one that more people have a hard time understanding. The technology does exactly what the name suggests — enables access to a mobile network through a realm of fixed wireless coverage based on an unlicensed frequency, such as Wi-Fi. In that sense, it's a cousin to IMS, a technology that layers previously separate vertical network technology silos into a unified convergent network architecture, an architecture which, among other applications, could enable a mobile/Wi-Fi convergent service. The two technology concepts also mistakenly have been perceived as competitors by some, though they are not necessarily competitive, nor even parallel concepts.

The two technologies, in fact, work their magic in completely different venues. While IMS serves to overhaul the traditional network core, UMA is purely an access methodology, one likely to be used in a residential deployment where Wi-Fi coverage already exists. UMA can and is expected to co-exist within IMS-based networks. “Technically speaking, UMA and IMS functionality can be together in the same network,” said Bengt Nordstrom, CEO of Northstream, the European subsidiary of U.S. research firm inCode Wireless.

“Part of the confusion is where the value proposition is, and to whom,” said Steve Shaw, director of marketing at Kineto Wireless, the company often credited with pioneering UMA and driving the creation of the carrier/vendor consortium that has campaigned for UMA to become a 3GPP standard, a mission that was accomplished earlier this year. “UMA is not competitive with IMS nor with SIP. It's the access solution that allows convergent services to happen. IMS is a core network philosophy.”

While it's true that IMS could render UMA or any other access medium somewhat less relevant, Northstream recently completed a comprehensive report on the UMA market that suggests several factors may be falling into place to support the immediate and longer-term viability of UMA.

The first of those factors is a growing desire among users for a fixed-mobile convergence (FMC) service. The growing abundance of Wi-Fi coverage, and the trend of wireless replacement of wireline service by consumers have hinted at the notion that an FMC service could be popular. Also, Northstream's report notes that a recent survey by Motorola of mobile subscribers in Western Europe showed that more than 50% of them would sign up for a UMA-based service within the next year if mobile calls made inside their homes were priced the same as wireline calls. Globally, the portion of total voice minutes that are represented by mobile voice service also is increasing. In the U.S., mobile minutes last year represented about 30% of total voice minutes; in the U.K., the number was 28% (Among the U.S. and several Western European countries, Portugal had the highest at 55%).

“Price may be the main criteria in the success of this kind of offering, but there are other indicators,” Nordstrom said. “Minutes are one of the indicators, but the viability of UMA also depends on the environment that you're looking at — there tends to be more wireless coverage inside residences in Europe and less in the U.S., where the properties tend to be larger and the population not as dense. More generally, a FMC service will be considered for anyplace that broadband Internet access is growing very fast.”

Although FMC services also could be considered new revenue-generating services, Nordstrom said he believes they will more likely develop as tools of customer retention in the broader competitive war between the various broadband service providers that are offering that access. Some of the early carrier support for UMA may bear out that view. For example, telco giant BT, which has no wireless network of its own, is offering its UMA-based Fusion service through a wireless MVNO with its own residential and enterprise customers as the primary target. Finnish operator Saunalahti is planning on doing the same thing.

A UMA-based FMC service could be an important ingredient in the residential broadband service battle because it takes the wireless aspect of a triple-play service package to another level.

“What other providers are doing with these solutions seems like it could be a response to what cable TV providers did in putting together the first triple-play services,” Nordstrom said. “I think cable providers might have an edge if they want to do UMA because they have the thickest broadband pipe into the home today.”

With competitive pressure increasing for all kinds of service providers, UMA is the most immediate option — and the only standardized one — for offering an FMC service that could promote some competitive differentiation from the pack.

“Right now, UMA is the only 3GPP standard allowing extension of mobile services over other kinds of fixed access infrastructures,” said Kineto's Shaw.

UMA actually has come quite a long way to become a 3GPP standard. It wasn't part of the 3GPP effort until June 2004, when the UMA consortium that had been driven by Kineto brought UMA to the attention of the 3GPP Technical Specification Group. It was approved as a standard in April 2005 under the title “Generic Standard to A/Gb interface, and is part of the 3GPP Release 6 document.

There currently is no other 3GPP standard that supports voice call roaming between different network types. Much of the IMS standard already has been approved in 3GPP Release 6, but some specifics of IMS remain to be included in 3GPP Release 7, including the Voice Call Continuity (VCC) specification, which would give IMS the mobile/Wi-Fi handoff capability similar to UMA that it now lacks. Nordstrom said the VCC specification probably will be approved next year, which would put it on a track for commercial implementation in about 2008.

But, many companies see the market for an FMC service developing sooner than 2008. BT Fusion is the only commercial UMA service so far, but several more are expected to launch within the next year. France Telecom will launch its Homezone UMA-based service next year, Saunalahti has confirmed plans for commercial launch and Teliasonera reportedly will begun launching UMA in 2006, in a rollout that will encompass the Nordic countries and the Baltic states over the next three years.

The immediacy of the market opportunity also has begun to stir action among hardware and software vendors that could contribute to UMA deployments. Early on in UMA's development, it was criticized as not having enough support from major infrastructure and handset vendors, but that is changing. Since last spring, several vendors have announced products based on UMA and developed partnerships to address the market opportunity.

Ericsson's unlicensed network controller, like Kineto's, is already available, and equipment giants Motorola and Alcatel also are coming to market with UMA gear. Meanwhile, Kineto recently forged a partnership with Nokia's infrastructure unit.

Many of the same vendors that are building UMA network gear also are some of the biggest names in the IMS standardization effort, though so far, there still are at least three notable vendors — Siemens, Nortel Networks and Lucent Technologies — that have not announced an equipment strategy for UMA.

“There are infrastructure vendors out there that aren't doing UMA yet, but they don't even have to because they can sell their equipment around someone else's UNC, like ours,” Shaw said.

For now, UMA-based UNCs are only available for GSM-based networks, but Nordstrom said it would be relatively easy to incorporate UMA technology in a CDMA network. “It would work the same way,” he said. “The UNC is basically like another base station controller in the network.”

Regarding handsets, earlier in 2005, only two small handset manufacturers had committed to make UMA devices, but more recently, Motorola, LG, Nokia and Samsung said they were producing handsets based on UMA technology.

“There will be deployment quantities of handsets in the first quarter or second quarter of 2006,” Shaw said.

Northstream's Nordstrom doesn't question the eventual availability of high quantities of UMA handsets, but he said it will more likely be late in 2006 before enough handsets are available to give carriers and their customers adequate choices.

“I think it looks promising,” Nordstrom said. “But you really need to have 10 or 15 types of handsets available, otherwise I don't think you have a full market.”

The near-term market prospects for UMA also are inspiring other kinds of companies to get involved. Reef Point Systems, a vendor of security gateways, revealed a UMA strategy earlier this year and just announced the product based on the strategic move — its UMA Security Gateway, according to Cam Cullen, Reef Point's vice president of product management.

The gateway encapsulates UMA calls and signaling in IPsec tunnels for secure IP transport. The gateway can support up to half a million simultaneous IP connections.

“When you get down to it, there aren't going to be a lot of people who want to adopt this if a carrier can't ensure some level of security,” Cullen said.

He added that while a security gateway functionality is called out in the UMA standard, there is not any more specific information about where or how it should be deployed by a service provider. So, companies like Reef Point are trying to fill the gap, and give carriers one less issue to worry about as they prepare to launch their UMA-based services.

“The security functionality could be deployed in something like a session border controller, but security is such a separate discipline that you risk affecting the scalability of this new service you're about to launch,” Cullen said.

Providers of handset-focused security solutions, such as Symantec, also are looking to support UMA deployments. “UMA traffic going through an IPsec tunnel helps, but any time you have a smart device that is roaming into one network from another, you are crossing administrative boundaries, and there are different kinds of policies and risks associated with that,” said Brian Hernacki, architect with Symantec Research Labs.

Ultimately, all of the companies getting involved with UMA will be watching some of the earliest trials and deployments for some indication of whether or not they are making the right decision.

“The belief that UMA has a brief window of opportunity is a very engineering-minded point of view,” Nordstrom said. “If it did, companies wouldn't deploy it; they would wait. But some of them believe they have a market to chase right now, and that UMA can help.”