Although femtocells have generated quite a bit of hype (read our report on a femto conference in 2007, for instance) for their potential to save money and boost call quality, analysts and manufacturers say that there is still much work to be done if femtocells are to have a major impact in the wireless market.

In addition to lowering the cost of producing femtocells, manufacturers need to figure out how to prioritise voice traffic over IP, to standardise femtocell architecture to avoid industry fragmentation and to mitigate interference issues with cell towers and other femtocells.

These may seem like daunting challenges to meet over a short period of time, says Paul Callahan, vice president of business development at femto maker Airvana, but they're ones that femtocell manufacturers will have to face in order to successfully bring their products to market.

"It's very hard to do right, especially when you're trying to do it right out of the chute," says Callahan, whose company specialises in broadband wireless infrastructure and has been developing technology for femtocells. "If it's anything like the way that security operated initially for Wi-Fi, then we're all going down."

The concept behind femtocells is easy to understand. Essentially, they're designed as small cellular access points that route nearby wireless voice traffic through pre-existing broadband connections. In this way, femtocells can provide VoIP for wireless handsets that can both improve call quality and save money by letting users make calls without using up their cell minutes.

Femtocell pricing

But one problem that could keep femtocells out of homes and offices is their price. Gartner analyst Akshay Sharma says that even though femtocells are still in their trial stages, many of the vendors he's spoken to say they're likely going to charge about $200 per femtocell unit. This is roughly two times the price that Vodafone CEO Arun Sarin says would be needed to make femtocells a widespread success.

Gartner analyst Deborah Kish thinks that $200 might be too much for a consumer to pay for a femtocell, but notes that it might entice enterprises that want to save money on their monthly telecom bills. As more and more enterprises try to save money by encouraging employees to use wireless handsets instead of landlines for business calls, they would be the ones to most benefit from having femtocells route call traffic through the broadband connection. Additionally, femtocells could boost the quality of mobile phones in the office to the point where employees could feel less inclined to rely upon their landlines for consistent service.

Of course, carriers could solve their pricing problems in the consumer market by subsidising losses in their femtocell divisions in order to keep prices low and entice early adopters. Currently, Sprint is offering trials of its Airave femtocell units for about $50, while charging customers a flat monthly rate of $15 for individual plans and $30 for family plans. And while vendors might be looking at charging high prices for femtocells now, notes Callahan, it's still far too early to tell what vendors will charge once the devices hit the competitive market.

"It's entirely an educated guess at this point," he says.

Standards and interference

Another challenge for femtocell deployment is the fact that no one has yet developed a single architecture that the industry can use as its standard. According to Sharma, there have been "a plethora" of models that have been designed based on SIP, UMA and other mobile architectures.

"There are around 13 different architectures that have been proposed at the Femto Forum," says Manish Singh, vice president of field engineering for Continuous Computing, a provider of femtocell software and integrated systems for telecom network infrastructure. "If this isn't resolved, it will translate to market fragmentation.

Interference is another issue that could hold up femtocells in the marketplace. Femtocells may work for users who live in rural areas and who don't live near cell towers, but users in urban environments could experience difficulty, especially if they live in apartment buildings with several femtocells clustered together in a tight area.

As Russell Cyr, vice president and chief marketing officer of BitWave Semiconductor, noted at a MassNetComms femtocell discussion panel this fall, being the first person on your block to have a femtocell might be akin to being the first person on your block to get a cable modem. In other words, both technologies will work splendidly until more people in your neighborhood adopt them, thus gobbling up more local spectrum and bandwidth, respectively.

Callahan says that this is a legitimate concern, and that femtocell manufacturers will have to develop smart radios for femtocells that will mitigate interference and adjust their signals based on their environments.

"In a multitenant-dwelling environment, you're going to need femtocells that adjust," he says. "Or if you have a femtocell that is close to a large cell tower, it will need to automatically adjust its power to overcome the tower's signal."

In the end, notes Sharma, the best way for femtocell manufacturers to find success is to avoid the crucial mistakes made by once-heralded technologies such as the now-forgotten VoDSL. According to a Gartner report written by Sharma and Kish earlier this month, VoDSL similarly suffered in the beginning from having too many competing architectural models, and faced questions about their end-to-end-management and QoS. And though Sharma is optimistic that femtocells will eventually be successful in the marketplace, he adds that VoDSL has demonstrated that no amount of industry hype can make a technology work efficiently enough to adopt.

"There was a lot of hype surrounding VoDSL," he says. "We've seen plenty of other technologies that have been overly hyped and have failed in the end."