Femtocell technology showing up in AT&T, Sprint, and Verizon plans

The tiny base stations may end up in Internet gateways, in businesses, and on the street, with adoptions by three major U.S. carriers

This week's CTIA Wireless conference may be something of a coming-out party for femtocells, the tiny cellular base stations that can send voice and data traffic onto inexpensive broadband links, but the devices eventually may look different from what most people expect.

The Femto Forum will sponsor a special showcase on the CTIA show floor in Las Vegas and host a roundtable on Wednesday afternoon. The industry group will also give details of a "plugfest" being held next week in southern France to prove interoperability among femtocell equipment from different vendors. The Forum expects more than 20 manufacturers to participate, testing out gear that was built to standards adopted last April by the 3GPP (Third-Generation Partnership Project). If the vendors can prove interoperability, that will be a step toward letting carriers and network equipment builders mix and match technology from many vendors, potentially driving costs down and femtocell adoption up, said Simon Saunders, chairman of the Femto Forum.

[ Keep up on the latest networking news with our Networking Report newsletter. | Stay ahead of advances in mobile technology with InfoWorld's Mobile Edge blog and Mobilize newsletter. ]

Also this week, AT&T is expected to say it will offer femtocells nationwide after conducting trials in several states beginning last September. Sprint Nextel's Airave device is already in full commercial deployment, and Verizon Wireless has said it plans to offer femtocells soon after it launches commercial LTE (Long-Term Evolution) mobile broadband service. Vodafone, NTT DoCoMo, and China Unicom also offer devices.

Though coverage complaints are as old as cellular technology, the recent explosion in mobile data use makes femtocells a timely innovation.

"For the iPhone users who are really frustrated over being dropped all the time, the femtocell may be the answer," said IDC analyst Godfrey Chua.

Though about 60 carriers around the world are conducting trials of femtocells, there have only been nine commercial deployments, according to the Femto Forum. This is for a technology that some observers thought would be widespread before the end of 2008. But vendors and service providers have had a hard time figuring out how to price and market dedicated household femtocells for consumers, as well as facing technical problems that were bigger than expected.

Now, some vendors and industry analysts say it may be other types of femtocells that ultimately deliver on the promise of the technology. These may include femtocells built in to consumer Wi-Fi routers, ones that boost coverage inside medium-sized enterprises, and outdoor units that form the leading edge of LTE deployments.

A femtocell is smaller than a microcell or picocell, two types of base stations that are often used to boost coverage in buildings. It is also different because it has a built-in RNC (radio network control) element. Usually, RNCs are located in the carrier's data center. Having an integrated one allows the femtocell to link in to any sort of wired broadband, such as a cable modem or DSL. The devices typically can cover about 5,000 square feet (465 square meters) and serve several mobile phone users in a household. The owner of the femtocell can create a "whitelist" of phones allowed on the device.

But making a femtocell talk to the handsets consumers already have, and with the carrier's back-end mechanisms for billing and other functions, has presented challenges.

"Getting femtocells to work with existing sets of mobile phones was a huge task -- much bigger than we expected," said Will Franks, founder and chief technology officer at Ubiquisys, which supplies femtocell technology to carriers, including Softbank in Japan and SFR in France.

One problem is that femtocells use features that are included in cellular base-station standards but not typically used in traditional cells, Franks said. As a result, phone makers don't design their phones to work with those features. For example, femtocells use "closed mode," which keeps unauthorized phones from accessing the cell. In development testing, Ubiquisys found handsets would hang, unable to jump on to any signal. The software on most phones couldn't be updated to add that feature, so Ubiquisys had to modify its femtocells to work around the shortcoming, he said. From now on, carriers certifying new handset models are testing them with femtocells, he said.

Femtocell maker IP.acccess, which confirmed it has been supplying technology for AT&T's trial deployment, has run into similar problems. For example, the "whitelist" of phones allowed on a femtocell is based on users' cell phone numbers and is supposed to update itself when a SIM card on one of the phones is replaced. IP.access found that femtocells didn't do that, so handsets would mysteriously stop working with the femtocell, said Andy Tiller, vice president of marketing. Once the problem was discovered, IP.access automated the process. It was one of many small issues.

"There are 101 of those things that need to be worked through," Tiller said. "It's a lot more complicated than people thought it was going to be. You have to retune everything."

Beyond the technical issues, service providers aiming their femtocells at consumers have to figure out how to sell a separate device that essentially delivers what the carrier has claimed to offer all along: adequate indoor coverage. There are many different possible models for selling femtocells. The problem is that offering the device points a finger at something negative -- the carrier's network capacity problem -- and asks the customer to do something to solve it. Indeed, research results announced this week by the Femto Forum showed that about 60 percent of the economic value of a femtocell lies in what the service provider can save on infrastructure.

"It doesn't sit right with a lot of people," said analyst Jack Gold of J. Gold Associates.

The offerings vary. In the United States, Sprint charges $99.99 for its Airave unit, plus an "Enhanced Coverage Charge" of $4.99 per month. Optional unlimited calling plans cost $10 per month for a single phone or $20 per month for multiple phones sharing the Airave. When AT&T began the first test offering of its Microcell device in Charlotte, N.C., last year, it offered a choice of two pricing plans. Subscribers could simply buy the Microcell for $150, or pay $20 per month for unlimited domestic phone calls made through the device. The latter plan included a $100 rebate on the cost of the Microcell.

Carriers are already having some success, said In-Stat analyst Allen Nogee. In-Stat estimates there were 380,000 femtocells shipped worldwide last year and will be about 2 million units delivered to customers this year. "I think it's gaining steam," Nogee said.

However, the real opportunity for in-home coverage lies with femtocells that customers don't even know are there, according to Nogee and other analysts. Two or three years down the road, the devices may be integrated into home broadband gateways or Wi-Fi routers, and the consumer will get better service without having to do or spend anything.

"For this to be successful, it has to align with a consumer electronics-like distribution model," said analyst Phil Marshall of Tolaga Research.

Carriers may also take aim at stores and medium-sized businesses, using slightly more powerful femtocells, observers say. Large enterprises and big public facilities such as shopping centers often get expanded coverage with distributed antennas, but this is usually too expensive for an organization with fewer than 250 employees, Saunders of the Femto Forum said. As with consumers, femtocells can make prized business customers more loyal to a particular service provider, he said.

The small-scale base stations may also play a key role in building LTE networks. Putting LTE femtocells in subscribers' homes could fill in enough coverage that a carrier could delay building some of its regular LTE cells, Saunders said. They could allow a carrier to delay the activation of 4 percent to 10 percent of the bigger cells, generating enough savings to cover the cost of household femtocells, he said.

Some vendors are also looking at LTE femtocells for public areas to fill in coverage or ease congestion. It would be relatively quick and easy to set up a femtocell on the street and plug it into existing broadband services instead of deploying conventional macro cell, said Franks at Ubiquisys. He estimates a cost-effective LTE version of these "metro" femtocells might hit the market in 2011 or 2012.

A new specification in the LTE standard, called SON (Self-Organizing Networks) should ease both of these types of deployments, said Steve Kemp, senior director of product marketing at Alcatel-Lucent. SON femtocells placed around a building or public area can find each other and optimize their own settings within that environment, saving expensive and time-consuming engineering work, Kemp said.

Though they have been a long time coming, femtocells could end up helping LTE reach users faster than today's networks did.

"When 3G was slowly being built out, they didn't have the benefit of femtocells to help fill in the coverage gaps," IDC's Chua said. Having them available from the dawn of the new network technology should make their inclusion that much easier.


Copyright © 2010 IDG Communications, Inc.