Merging Wi-Fi to cellular phones is still the hot idea of the year, but the technology so far put forward by operators and telecoms vendors could find itself looking lame compared with a more IP-centric approach from a different direction, that may be here by the end of the year.
Devices that can handle both Wi-Fi and cell networks are painfully emerging. Services like Vonage (which just launched in the UK) have convinced us that Voice on Wi-Fi is not such a bad idea (though we still have our reservations).
What we are waiting for the infrastructure - and the commercial model - to let us handle voice calls across both media.
A UMA proposal
Airespace, the wireless LAN switch start-up recently bought by Cisco, has announced that its is working with cell-Wi-Fi convergence specialist Kineto Wireless, to allow businesses to use mobile phones over their office Wi-Fi networks, and claims to have already managed to hand calls over from one mode to the other, in 40 to 80 milliseconds.
However, the technology basis is the UMA (unlicensed mobile access) specifications - of which Kineto appears to be a key enabler - a set of specifications which proposes convergence entirely on terms to suit the mobile industry's legacy.
UMA's promise, for instance in the Bluephone project which BT has promiosed for eighteen months and may yet actually deliver (see Bluephone's Wi-Fi future revealed), is to allow mobile operators to use our home and business WLANs, to deliver service indoors - and charge us for the privilege. It doesn't use SIP, and it doesn't use the Internet in a real way, because actually, it's all about bolting Wi-Fi onto legacy systems.
IT managers want something completely different. They want to move calls off the mobile network, and save money, when they reach WLAN coverage.
UMA people make big claims, but observers have doubts. "By working with Kineto to ensure seamless interoperability with UMA technology,", says Alan Cohen, vice president of marketing at Airespace, "we are paving the way for widespread Voice over WLAN (VoWLAN) market adoption."
Dean Bubley of Disruptive Analysis snorts: "I'm starting a campaign against the word 'seamless'. We need seams. Without them our trousers would fall apart."
At this point, a visible join between the two networks is no bad thing, if it helps managers track costs. Seamlessness is no good if it comes at the cost of any control over call costs.
Signs of hope?
Another alliance appeared rather quietly at 3GSM, which promises to move calls between cellular and Wi-Fi, but uses SIP, and makes the most of the cheaper routing available on the Internet.
The MobileIgnite alliance, based around BridgePort's NomadicOne convergence gateway, is offering operators a system in which users get a SIP address on their phone, and a "presence" function that tells the system whether they are on the cellular network or on the Internet (wired or wireless), or have changed devices.
It bridges from SIP to the SS7 signalling protocol used in all fixed and mobile telephone networks. It works fine now, and will work even better when those networks migrate to the IMS (IP Multimedia Subsystem) specified by the ever-productive 3GPP standards group, according to Sanjay Jhawar, vice president of marketing at Bridgeport.
Bridgeport had a demonstration at 3GSM, using servers from IBM, specially built client software from PCTel, cellular service from operator WorldTel, and Wi-Fi networks from Airespace (all MobileIgnite members). Users could make and receive calls at the same SIP identity, using either the Wi-Fi or cellular function of an HP iPaq 6315 (there's more on Brigeport's press release).
The service, as demonstrated, takes SIP to mobiles. It allows a user to make calls from their mobile, which are delivered over the IP network as a standards-compliant VoIP call. It also allows calls for someone's VoIP number to be forwarded across the mobile network, if the presence function shows there is no VoIP connection to the person. "This is not a call-forward," says Jhawar. "That would be circuit switched, with a mobile termination rate. It goes as a SIP call."
At 3GSM, where the mobile networks were heavily loaded, the ability to make and receive calls over Wi-Fi was apparently very useful.
So when can we buy it?
Bridgeport has yet to sign up any operator to deliver the service to end users, but Jhawar believes the demand will be there, with products coming to end users by the end of the year: "It is up to the service providers. They have to figure out what group they are selling to, and what phones are appropriate."
The politics may be tougher than that: "Offloading calls to VOIP has its attractions, but operators are lukewarm on subsidising handsets that could be used to bypass their networks," said Unstrung Insider chief analyst Gabriel Brown in a recent report.
However, operators may be forced into a technology like this, when they are put under pressure by services that bypass them altogether, like Vonage. Services will appear first in North America and the Far East, where operators are under most pressure from alternatives, said Jhawar, but they could appear in Europe by the end of this year, he said. The handsets may be the biggest problem, although
Prospects could be good
The two approaches don't conflict - witness Airespace's presence in both camps - but it's fairly obvious that UMA-based services are going to be pretty uninteresting when IMS-based services (or those which, like MobileIgnite, anticipate IMS) are available.
Systems tailored for legacy networks - like UMA - might have a short window in the market, if networks upgrade quickly to IMS. And it looks as if they might - already, although IMS was designed by 3GPP for mobile networks, it is being used by BT in its own core upgrade.
"UMA is not fully fledged IP," says Jhawar. "The approach might make some sense in short term, but the new cost structure demands pure VoIP. UMA will be deployed, but migration to IMS will be a stronger factor."
However popular UMA is, network managers should be relieved to know that something more forward looking is on the wya.