Taking the Wrong Train to FMC?Taking the Wrong Train to FMC?

To really penetrate the enterprise market, the cellular carriers are going to have to get over the idea that they can just sell consumer services to businesses.

Last week, WLAN switch maker Meru Networks announced a joint agreement with T-Mobile to market the latter's UMA-based fixed-mobile convergence (FMC) solution to business customers. T-Mobile has joined Meru's Wireless Interoperability and Network Solutions (WINS) partner program, and tested the ability of their UMA-based handsets to handoff connections to a Meru enterprise WLAN infrastructure.The participation of WLAN switch vendors is a new twist on the FMC market. Aruba Networks had privately disclosed a similar UMA capability (also using T-Mobile's service) some months back, but we have not seen any reports of actual customer deployments. As a general rule, the WLAN companies, like the IP PBX vendors, depend on third-party adjuncts like Agito, DiVitas, and Varaha to address dual mode FMC applications.

For those who are not familiar with it, UMA is the dual mode Wi-Fi/cellular FMC solution used in T-Mobile's consumer-oriented Unlimited Hotspot Calling service (formerly called "[email protected]"). When a UMA user comes within range of a predefined wireless LAN, in-progress calls are handed off to the WLAN and the cellular billing stops. The customer's Wi-Fi router is connected to a broadband Internet connection that forwards packets of GSM voice to a UMA controller connected to T-Mobile's network. The consumer version of the service costs $10 per month (on top of a premium cellular plan), and all calls made or received over the Wi-Fi connection do not use plan minutes. There was no indication whether that same $10 per month price would apply to enterprise customers.

One appealing feature of UMA for WLAN vendors is they don't have to do much to support it. When a UMA device is operating over the Wi-Fi network, it simply takes packets of encrypted GSM voice and signaling, sticks them into Wi-Fi frames, and sends them. Essentially all you have to do is store the SSID (i.e. network name) of the corporate WLAN in the handset, and it will roam onto that WLAN switch network the same way it roams to a home Wi-Fi network. Once in the WLAN switch environment, the user device can roam from AP to AP and the call will be handed off transparently. Meru has built its reputation largely on its abilities in the VoWLAN area with a single channel RF layout, zero delay hand-off, and proprietary scheduling mechanism that can support a greater number of simultaneous voice calls than competitive WLAN switch offerings.

However, cost reduction is only one goal in FMC. Equally important are the benefits of one number access, single voicemail, maintaining control of the telephone number, and extending unified communications (UC) functionality to mobile users. By itself, UMA does not merge the wired and wireless experience; in fact, a UMA solution operates completely independently of the PBX. The result is that every UMA user would have two phone numbers: an office number and a UMA cellular number.

This might seem like a rather gross oversight, but the whole idea of UMA is based on the requirements of consumers, particularly those who have "cut the cord," abandoned the wired telephone network and get by with a cell phone. Few enterprise users fit that mold; the vast majority of business users also have a desk and require a business number. Of course the cellular carriers never made much of an attempt to address the requirements of enterprise users in the first place. From a business perspective, if we have business contacts calling our employees on a cellular number, that cellular number really should be "owned" by the business. Otherwise we run the risk of having those contacts following the employee when he or she leaves the company.

Getting what businesses need from UMA involves adding some more elements to the network. The minimum capability a business will require is simultaneous ring (a.k.a. "Extension-to-Cellular"). That function can be provided either through the PBX or with an adjunct server like the BlackBerry MVS/Ascendent or the CounterPath Enterprise Mobility Gateway. With simultaneous ring, incoming calls to the desk number (i.e. "business number") are rung to the user's desk set and simultaneously rung to their cellular number; the user can answer the call on either. The key advantage is that there is one number to reach that party, and the cellular number is effectively hidden from the caller (Note: Truly hiding the cellular number also involves routing outbound mobile calls through the PBX).

The good news is that if the cellular number is a UMA number, the cellular carrier will know whether to route their end of that incoming simultaneous ring call over the traditional cellular network (i.e. the radio access network) or through the Internet/Wi-Fi path. Of course the network diagram winds up looking like you dropped a fistful of spaghetti onto a white tablecloth!

Providing UC capabilities to that mobile device raises the bar yet another notch. For mobile UC, we use the cellular data service to deliver presence-enabled directory, visual voicemail, caller ID info and a host of other UC features to the mobile device. That means we will need a business-capable smartphone that is also UMA compatible. Most UMA handsets, including the new HTC Shadow and Nokia 7510, lack a QWERTY keyboard; not surprisingly, they are consumer handsets. RIM does offer some UMA-capable BlackBerry models (i.e. the ones with model numbers ending in "20" like the Pearl 8220 flip phone). Of course, a mobile UC capability that depends on cellular data service is not playing to T-Mobile's strong suit, as they are the last of the national cellular carriers to roll out 3G services.

The bigger problem with the whole UMA approach is that it has only one proponent in the US, T-Mobile. Number four among the national cellular carriers, T-Mobile's major accomplishment is that they're not doing quite as badly as Sprint. Fourth quarter results indicate that T-Mobile had 621,000 new additions, 1/3 fewer than the prior year. Further, 57% of those additions were the less attractive non-contract customers, and churn increased from 2.8% to 3.3%. Sprint won't announce their fourth quarter results until February 19, but industry estimates are that they will lose 1.1 to 1.3 million subscribers. When we compare Verizon and AT&T, who gained 1.0 and 2.1 million subs respectively in the fourth quarter, we recognize an environment where the rich get richer and the poor get pink slips.

UMA is gaining some traction in Europe, where Telia Sonera is offering the service in Denmark and Sweden, Ono in Spain, NetCom in Norway, Saunalahti in Finland, and Orange in France, Holland, Poland, Spain, and the UK. However, with recent announcements of femtocell offerings from both Verizon and AT&T, it is fairly clear that the major US carriers are not going down the UMA path (Sprint introduced their Airave femtocell offering in 2008).

It is commendable that T-Mobile is at least making some attempt at offering an FMC service, but we need something that is designed for the enterprise, supports a range of handsets and is in step with the overall move to unified communications. T-Mobile does offer the BlackBerry MVS solution, but that works in conjunction with a BlackBerry Enterprise Server and supports RIM devices primarily. To really penetrate the enterprise market, the cellular carriers are going to have to get over the idea that they can just sell consumer services to businesses.To really penetrate the enterprise market, the cellular carriers are going to have to get over the idea that they can just sell consumer services to businesses.