The three recent FMC launches provide real-world illustrations of two different technology approaches that have been debated since carriers first started taking FMC seriously about five years ago. Embarq has chosen a solution based on session initiation protocol (SIP) that borrows some concepts from the emerging IP multimedia subsystem (IMS) standards. Many carriers have begun to deploy or have plans to deploy IMS, which will support appealing new services by enabling different devices and networks to share functionality.

Cincinnati Bell and T-Mobile, on the other hand, have opted to use unlicensed mobile access (UMA), an alternative specific to GSM wireless networks that makes a Wi-Fi node appear like a cell site to the cellular network. Wi-Fi traffic is tunneled through the Internet to a gateway that connects it to the cellular network.

Whichever solution a carrier chooses, a key function of the FMC infrastructure in a dual-mode handset deployment is to control handoffs between Wi-Fi and cellular networks. There are some key differences between how UMA and SIP solutions treat handoffs that have important implications for network operators.

Craig Gosselin, chief marketing officer for NewStep, which has based its FMC solution on SIP and IMS, explained what happens in that environment. “Through signaling, you can create new sessions and drop old ones,” he said. The handset communicates with network-based software, which Gosselin likens to a switchboard that can establish a connection to the new network. “Our software and technology makes sure the new leg is working, then breaks the original connection,” he said.

While that approach may be fine for a carrier such as Embarq that is primarily focused on the wireline market, it may not be the best solution for a wireless carrier, said Shaw of Kineto, which has based its FMC solution on UMA. The reason, he said, is that when an IMS solution is used, wireless network operators lose control of the call when it migrates to the Wi-Fi network and therefore have no idea how extensively the customer is using Wi-Fi.

“UMA is a technology for mobile operators,” Shaw said. “It connects like a regular network element. The network doesn't know you're running over IP, Wi-Fi or broadband. It thinks it's connected to a cell tower.” Software in the network, however, can recognize when a call comes in from a Wi-Fi connection and can bill — or, more likely, not bill — accordingly.

But Gosselin said UMA is too limiting. “Our partners are saying they need something very flexible,” he said. “They need it to work in an IMS network, and with CDMA or GSM, regardless of media type.”

Pre-IMS, SIP-based solutions rely on an emerging standard called voice call control (VCC) to handle voice handoffs between Wi-Fi and CDMA or GSM. But as carriers migrate to IMS, the same solutions should be able to handle other media types so that, for example, video content designed for delivery over a broadband connection could be shifted to a mobile device, Gosselin said.

Shaw challenged the notion UMA is not compatible with IMS. “Because UMA is at the transport layer, it doesn't know what it's carrying,” he said. Assuming the UMA infrastructure was connected to an IMS server, he said, “if you were rolling out applications based on IMS, those applications would run over UMA.”

UMA can work with IMS, but only so far, said Luke Thomas, a program manager for research firm Frost. “It only supports existing mobile operator services designed for GPRS or the data channel, rather than working at the session layer between service domains,” he said.

However, Thomas added that such a limitation actually could be appealing to mobile operators. If IMS is opened up to its full set of capabilities, enabling it to link service domains, mobile customers could use it to detect a friend's presence on a low-cost or free voice-over-IP network such as Skype. The customer could then use that network to contact the person, eroding the traditional network operator's minutes of use. Potentially, UMA could provide a means of preventing that scenario, Thomas said. He cautioned, however, that enhancing UMA to support SIP would require major changes to some GPRS protocols that might not be acceptable to some mobile operators.

Integrated carriers that have nationwide wireless and wired networks have fewer compelling reasons to deploy dual-mode offerings than less diversified carriers such as Embarq, Cincinnati Bell or T-Mobile — at least in the near term. But if the early dual-mode services are successful, integrated carriers could feel pressure to respond with converged offerings of their own.

“T-Mobile could be a game-changer,” said Dekkers Davidson, a director in the converged media practice for consulting firm Oliver Wyman. “To the extent they can solve the quality problem in the house, over time it becomes more and more compelling.”

On the commercial side, NewStep's Gosselin expects converged offerings to increase in popularity as businesses recognize productivity benefits, which will become more important than potential cost savings.

However, dual-mode offerings face a number of market challenges, Thomas said. For example, only a few manufacturers offer dual-mode handsets today, and some may lack features that users have come to expect, such as a built-in camera or MP3 player. The requirement to use a specific wireless carrier also may be an impediment, particularly in the business market, where numerous employees might need to switch carriers.

Dual-mode handsets also are not the only solution to the problems they have been designed to address. For some business users, one-button call-switching options such as Embarq's Smart Connect service may be a simpler and more cost-effective way to reduce reliance on cell phones in the office and may provide an equivalent productivity boost. One-number services that simultaneously ring a customer's cell phone and home or office phone also can help minimize cellular minute usage and address poor in-home cellular coverage.

Sprint's recently launched Airave offering also could be a strong opponent. Airave uses femtocells, which are essentially mini base stations installed in a customer's home in combination with a broadband connection to enable customers to use their cellular phones in their homes without per-minute charges. That solution, which is cost-competitive with the T-Mobile and Cincinnati Bell dual-mode offerings, could minimize cellular minute usage, address spotty cellular coverage and potentially eliminate the need for a wireline phone.

FMC developers, meanwhile, are looking toward their next generation of dual-mode features. “The biggest thing you'll see is 3G handsets,” Shaw said. “If operators are going to 3G, it's because they have great multimedia applications and feature-rich services.”

Such high-bandwidth services are ideally suited for high-bandwidth Wi-Fi links and will drive interest in converged 3G/Wi-Fi handsets, he said. By off-loading backhaul traffic to a broadband connection, Shaw said, operators also will minimize their need to increase the capacity on their backhaul networks.

Kevin Fitchard contributed material to this report.