Femto cells vendors are talking up a storm at the moment, about the tiny 3G base stations that can boost network coverage inside consumers' houses. In the next few months, they will have to move on to more than talk - delivering new and more complex applications, and agreeing on standards and business models.
It will get more intense - because the competition is about to start in earnest, said Andy Tiller, vice president of marketing at femto vendor ip.Access, speaking to us after Cisco invested in the company last week.
At this stage, potential customers haven't seen real femto cells. The only commercial market trial of a femtocell is Sprint's one in the US, and that isn't a 3G device, but a CDMA one. It has a very basic proposition - cheaper calls in the house - that operators can already offer with other technology.
New services required
"Operators are telling us that is a good start, but they would like new propositions that would only work at home," says Tiller.
These femto zone services might include automatic texts when one of the residents' phones comes in range (so people can tell if their family are at home), or linking to the home network, so all music, photo and video content on the phone gets synched with the home server automatically in the background, showing the photos on the home TV, perhaps.
Tiller is bound to be interested, because ip.Access' new investor, Cisco is vitally interested in the "connected home" on behalf of its subsidiaries Scientific Atlanta and Linksys, who make set-top boxes and home gateways.
"There are a lot of companies working in femtocells," says Tiller. "It's going to be quite a competitive area."
What about standards?
What if there is such a rush on, that standards get overlooked? "Operators don't want to be locked into one vendor's solution," says Tiller. "What they really want is for multiple vendors at both ends. On the other hand, they are so keen to get femtos out quickly, they don't want to wait for standards to be defined."
That makes the Femto Forum's programme to agree on femto architectures, announced today, really important. "The Femto Forum is taking a lead in getting everyone together," says Tiller. "I've been really impressed with the amount of serious intent from all vendors [in the Forum] to find a way to give the vendors what they want. We recognise that if some of us succeed in locking operators in, it’s not going to be good for the industry."
The Forum lets vendors get to know their rivals: "While fiercely competing on the outside, we have to collaborate," says Tiller.
The first phase of the Forum has been to map out the problem, describing a generic network architecture. "There are actually more approaches than vendors," says Tiller, since each vendor is planning to serve multiple operators, and different business models. Now a new phase will try to harmonise a few options, through consensus. "I sense a genuine desire from vendors to make that happen."
The femto business model will determine exactly what features will be built into the femto device. But that model is still being defined - and will continue to evolve till commercial trials have taken place.
Despite this, there's already a good understanding of what the possibilities are, says Tiller, with basic features well understood, like offering cheaper calls, and keeping the user informed whether the call is on the femto or not, so they don't run up high mobile bills because the femto is switched off.
The business model may determine what kind of access control the femtos will have, and whether they allow devices on the macro network to use them. "It's harder to sort out a policy on the ability to handover from the macro network to the femto," says Tiller. "Some operators require that, others don't."
Defining a protocol
Whatever the features, vendors will have to standardise the way the femtos connect to the core network, so operators can adopt femtos from multiple vendors to back end aggregators. Those connections will have support a standard set of features. "The actual protocol by which the femto talks to the core network is done differently by different vendors," he says. "That's the bit we need to fix in order to enable interoperability."
Without this protocol, a femto would be as much use as a mobile phone without a GSM protocol, he says "Just as interoperability based on standards was the key to GSM's success, we need this for the market to take off."
"Operators are telling us, it's fine," he says. They're ready to do a certain amount with the proprietary systems available today, but they won’t go beyond a certain point unless they can source from multiple vendors."
The Femto Forum isn't a standards making body, and any femto standards will have to be ultimately agreed in bodies like 3GPP, but the Forum can do a useful job getting the process off to a good start, he says: "There can be a lot achieved by consensus, before formal standards."
Among other things, he says, the Forum brings in real world experiences, and provides a reality check. For instance, the 3GPP already has a work item on femto cells, which includes a suggestion that femto cells should - like macro cells - be able to handle calls from handsets moving at 30mph. Input from the Femto Forum should be able to point out unnecessary suggestions like this, says Tiller. "The Forum will provide a means for the standards process to be better informed."
Femtos versus Wi-Fi
Like everyone else, he says femtos are complementary to Wi-Fi, the alternative way to handle voice on a local base station: "Will femtos replace Wi-Fi in the home? Of course not. Wi-Fi is established, and lots of devices will use it."
But 3G-based femtos are the "more natural" approach for mobile phones, he says, because "the phone doesn't have to do anything differently from outdoors, it all happens for free, and is available in all phones and it solves problems like battery life. It's more natural for a phone to carry on using the same air interface."
Also, femtos actually make mobile batteries last longer, which , he points out is "the opposite to Wi-Fi."
Wi-Fi might seem to have a head start. There are dual-mode phones in the shops today, but no femto cells. But Tiller disagrees: "The race is much more even than some would imply. Wi-Fi will just be starting on the phones people want to buy when femtos arrive."
Solutions like dual mode UMA are available on a limited set of handsets today, he says - "and they're all boring handsets." The UMA standard won't even be available on 3G phones till later this year, he says, while femtos work with all 3G phones from the start.
The exception, he acknowledges, is the enterprise, where dual-mode handsets could be part of a package, and hook up to existing IP PBXs. "In principle, Wi-Fi has more of a chance of seeing off enterprise pico cells, than it does against consumer femto cells." Even there, though, he thinks enterprises may be less willing to "tune" their Wi-Fi for mobile voice, if there's an off-the-shelf solution available, which operators will pitch hard.
So, over the next few months, watch for standards, watch for public trials, and watch for new alliances, as the femto market attempts to get real.