Fixed/Mobile Convergence Technology and Markets

Most VoWLAN phones sold today are WiFi only rather than cell phones with WiFi radios. But analysts are unanimous in expecting over 100 million cell phones sold in 2010 to have WiFi built in, outselling WiFi-only phones by about an order of magnitude. There is also a consensus among analysts that WiFi's primary use in cell phones will be for VoIP. These phones will fall into two categories: low end phones and smart phones.

Whereas the technical challenges of VoWLAN have been substantially solved, VoWLAN cell phones still present usability challenges to customers and business challenges to Mobile Network Operators (MNOs) and cell phone manufacturers (OEMs). I will look at the usability challenges in a later column.

In the USA, wired Internet access is all-you-can eat; once you have Internet service, all peer-to-peer applications, including voice, are effectively free. This sounds great to you and me, but it is a nightmare to service providers who derive most of their revenues from voice service. Of course you do pay extra for some types of Voice over IP, like Vonage, and some types of WiFi are billed as a service, like hot-spots. These don't fall into the category of nightmare to the MNO. And MNOs wouldn't see VoWiFi as a nightmare if they could bill for it. In that case it would be attractive, because the WiFi would offload the cellular network, enabling more capacity without expensive build-out, and it would improve coverage in residential areas, which currently is often poor. MNOs have two candidate technologies for charging for VoWiFi: UMA (Unlicensed Mobile Access) and IMS-SIP (IP Multimedia Subsystem - Session Initiation Protocol). UMA is the short-term solution, because it tunnels regular GSM signaling and voice through the Internet. IMS is the long term solution because it adds to the Internet a mechanism for each packet to be tracked and potentially billed.

But what about applications like Skype? And what about corporations that decide to peer or federate their IP PBXs so that calls between them don't pass through a voice service provider; what about enterprise mobility controller solutions that tunnel voice over VPN connections to employees worldwide? None of these are billable by UMA or IMS. These applications are a natural result of seeing the phone as a tiny PC, a platform on which you can load any application, and which you can connect to the Internet with whatever NIC happens to be convenient (USB, WiFi, HSPA, WiMAX or whatever).

This is where the distinction between low-end phones and smart phones comes in. The low-end phones will do WiFi voice with UMA or IMS, and users will be billed for it. This will also happen with smart phones, but smart phones can potentially also have the option for WiFi voice that is not billed by the MNO.

MNOs tend not to like this kind of open platform phone - why would they, when it has the potential to erode their revenues? But phone manufacturers don't have the same aversion. For example, Nokia is eager to promote the idea of the phone as a substitute for the computer, with two product lines: the Nseries, positioned as a consumer-grade multimedia computer and the Eseries, positioned as a business- grade device.

So users and OEMs like open-platform smart phones with WiFi connectivity, and MNOs are suspicious of them. In a world where handsets are subsidized by MNOs, and increasingly branded by them, the MNOs have the whip hand over the OEMs. As a result, it's conceivable that open-platform smart phones will continue to be a niche. For example, look at Apple, who one might expect to champion the open platform model. Although it runs a version of OSX, the iPhone is not an open platform.

Steve Jobs' excuse for this, "Cingular doesn't want to see their West Coast network go down because some application messed up," echoes a disingenuous rationale advanced by some service providers. Bugs in client protocol stacks have caused cellular network outages in the past, but those bugs were in network-layer software in closed phones so this argument is irrelevant to an application ban. Modern networks are designed to withstand attacks far more sinister than an application messing up. If Cingular's didn't fall into that category it would be going down no matter what restrictions they put on their phones, especially since they are eager to connect increasing numbers of wide-open PCs with HSDPA cards.

MNOs are not the only service providers in this game. VoIP (define - news - alert) has already brought MSOs (Multi- Service Operators, aka cable companies) into the voice service game, and they will most likely make some kind of MVNO (Mobile Virtual Network Operator) move to turn their "triple play" into a "quadruple play." And there are other MVNOs focused exclusively on FMC, notably Sotto, which sees Enterprise FMC as a more attractive opportunity than consumer FMC.

Michael Stanford has been an entrepreneur and strategist in Voiceover- IP for over a decade. In his current consulting practice, Michael specializes in VoIP wireless networks, both WiFi and WiMAX. You can reach him at [email protected]