The magic of the Microcell

There’s something different about AT&T’s new Microcell from the other femtocells being offered in the U.S. by Verizon Wireless (NYSE:VZ, NYSE:VOD) and Sprint (NYSE:S). Not only does the Microcell support 3G data, but AT&T (NYSE:T) isn’t charging its customers to use it, which could make it the ultimate extension of AT&T’s dual-network strategy to make wireless data ubiquitous without straining the capabilities of its high-speed packet access network.

By using a femtocell, a data connection bypasses the two most congested parts of the wireless network: the radio access network — which not only has built-in spectrum, deployment and maintenance costs, but is inherently limited by the amount of spectrum an operator owns — and backhaul transport, which is dependent on expensive fiber links to scale. Instead, the femtocell uses a customer’s own home or business broadband connection to tunnel that traffic over the public Internet directly to the network core. And as femtocell technologies improve, the majority of traffic that is bound for the Internet can be offloaded at the femtocell itself, bypassing the operator’s network entirely.
Unlike its voice counterpart, data is primarily consumed indoors, so a grid of femtocells at homes and offices could ease enormous amounts of traffic from the wide-area macro network. That in turn could save an operator considerable sums in network expansion, operations and spectrum acquisition costs. In fact, AT&T maintains that if network data traffic continues to grow at its current pace, no single operator can meet those demands with a 3G or 4G network alone. AT&T declined to discuss its microcell strategy for this article, saying that the service is still only in the trial phases, but several AT&T officials have made it clear that AT&T isn’t counting on HSPA or even its future long-term evolution network to shoulder its entire mobile data burden.

“I don’t think a single macro wireless network is sustainable over time, given the pace of spectrum availability and what’s actually out there in terms of fixed spectrum,” said John Stankey, president and CEO of AT&T Operations at his SUPERCOMM keynote in October. “I think the reality is that there will be macro and micro networks. We’re starting down that path at AT&T, focusing on how we make Wi-Fi and licensed spectrum a more seamless experience for customers. Other variants will come up on this over time. But to manage the kind of mobile experience customers will require and truly have the quality of service they need will require a multiple approach to network management in the wireless space. You’re going to see micro and macro in terms of licensed and unlicensed spectrum. This is a key architecture element we’re all going to have to come to grips with.”

Ralph de la Vega, CEO of AT&T Mobility, made comments in the same vein in an interview with TelephonyOnline: “There will always be a mixing of fiber and spectrum. For example, as we deploy our U-verse technology, we’ll be taking high-bandwidth services to the home. When you’re in your home, there’s no reason to use a wireless network. You have a broadband connection at the home you can use with Wi-Fi. There are ways to offload networks using Wi-Fi and therefore minimize the need for additional spectrum or additional bandwidth.

“You’re definitely going to see more femtocells or Wi-Fi that will allow traffic to be sent through a fixed network rather than just ending everything over wireless. That jury is still out. In the end, it’s not going to be one thing; it will be a combination of things: taking fiber closer to the home, Wi-Fi and femtocells. A combination of all of those is going to help us manage bandwidth and provide a great experience to our customers no matter where they connect.”

Both Stankey and de la Vega’s emphasis on Wi-Fi was no coincidence. AT&T has already become highly dependent on its 20,000-hotspot Wi-Fi network to augment its 3G network as more and more of its devices — the Apple iPhone being the most prominent example — come with Wi-Fi capabilities. Last quarter, AT&T saw connections from smartphones and other embedded devices surpass PCs on its hotspot networks. Given that AT&T smartphone, U-verse and DSL customers can access those hotspots, the Wi-Fi network is gradually transitioning from a stand-alone source of access revenues to one whose primary function is to support AT&T’s other data networks.

So far AT&T hasn’t indicated it will take the same approach with femtocells. Its trials actually offer similar services to that of its competitors, Sprint and Verizon Wireless, which haven’t been as aggressive in finding ways to offload traffic from their 3G networks. AT&T is selling the Cisco Systems-designed (NASDAQ:CISCO) Microcell (with software from IP.Access and silicon from PicoChip) as a stand-alone device or packaged with a $20 unlimited calling plan within the femto zone. Like Verizon, AT&T doesn’t charge customers a monthly service fee to use it. And while AT&T’s femtocell supports data, customers only get unlimited data usage if they already have an unlimited data subscription. If AT&T is planning on using the femtocell as means to unburden its 3G network, then it needs to present a compelling reason to its customers to buy it. So far, AT&T hasn’t supplied that reason, said Peter Jarich, senior wireless infrastructure analyst with Current Analysis.

“This has to have something with more meaning to the customer,” Jarich said. “Everyone wants a better signal in their home, but most people think it’s the responsibility of the carrier to provide that coverage, not their responsibility. Unlimited calling plans are nice, but most subscribers have such huge buckets of minutes, there’s not much more value they can get out of a femto voice plan. You have to offer something more than just capacity and coverage to get people to buy these things.”

Jarich thinks AT&T has an opportunity to offer personalized applications through the femtocell, turning it from a mere coverage enhancer into a service that a customer would actively seek out. One of the most constantly cited applications is family presence notification: Whenever a child leaves the home, bringing his or phone outside of the femto zone, a text message is immediately sent to the parents, allowing them to track their children’s coming and goings. A much more integrated app could turn the phone into a remote control for the home. A smartphone screen, for instance, could change from its normal view into an home appliance interface as soon as a customer steps into the femto zone, allowing a customer to channel surf the TV, play voicemail through the stereo or turn the security system on and off.

The Femto Forum has been widely promoting the idea of femto apps, going so far as to promote a standard set of application programming interfaces to help developers create applications across different vendors’ femtocell platforms. Following such a model, femtocells could become a device application platform much the same way Android or Symbian have become smartphone application platforms.

“When we get to these innovative services, the uniqueness of the femtocell really comes out,” said Andy Germano, vice-chairman of the Femto Forum. “It’s not just about unlimited calling plans anymore.”

By turning the femtocell into an applications platform, a operator like AT&T can do something it can’t accomplish with Wi-Fi, Germano said. It could build what is essentially a private data work complete with sticky personalized applications that would largely be subsidized by the customer, who would not only pay for the femtocell itself, but the monthly service fees and provide the backhaul link. What’s more, if an operator were to reach a critical mass of femtocells on the 3G network, it could fundamentally change the way it designs its future 4G network, Germano said.

“An LTE deployment could match cost to revenues,” Germano said. Instead of building a macro network everywhere, operators could build it in metropolitan cores and heavily trafficked corridors. It could then fill in the gaps with femtocells, knowing the 3G network will kick in if a customer happens to be outside of both macro and femto coverage, Germano said. “Rather than take the attitude that you build and they will come, you can build as they come,” he added. “Rather than go from big to small, they can go from small to big.”