One of the biggest challenges facing LTE is the problem of supporting legacy voice, and despite separate initiatives led by T-Mobile and Nokia Siemens, this is unlikely to be solved in the coming year. Like WiMAX, LTE services are expected to be data oriented in the early years, but while this is fine for new operators, the 2G and 3G players have to find an effective way to support voice, which remains a huge cash cow, and to integrate their old and new networks. Meanwhile, they face a rising challenge from the voice services from major portals like Google.
When LTE was first conceived, two assumptions were made that have proved to be false ‘ that by the time it rolled out, just about all carriers would have migrated to full IP and IMS (IP Multimedia Subsystem); and that voice would no longer be a major source of competitive edge. In fact, the move to IMS has been slow and piecemeal, and most cellcos need to support the ‘missing link’ between traditional cellular networks and full IP, for many years to come. This ‘missing link’ is proving the technology challenge for voice in the 4G environment, yet operators cannot just ignore voice. This may be a declining part of their total revenue mix, but it is still typically three-quarters of revenues, and still the most-used user application.
And a key growth strategy for many cellcos is luring landline customers onto the wireless network ‘ which makes control of voice activity essential if the carrier is not to end up merely providing a bitpipe for third party IP telephony providers like Google Voice, Skype and soon Vonage. Plus the voice market, while declining as a percentage of revenues, is still worth $692.6bn a year, so is not one to be thrown away. But margins are low, so operators will aim to use existing networks as much as possible, especially as circuit switching remains more efficient than VoIP for delivering voice ‘ rather than investing in IMS, the main way to support carrier class VoIP in LTE.
VCC (Voice Call Continuity) is the primary method of handing off voice calls from LTE and HSPA+ to older networks but is unlikely to be fully supported in 3GPP standards until release 10. And most carriers will not implement IMS with LTE from day one, also limiting the ability to carry voice efficiently on the new network. By contrast, WiMAX had VoIP capabilities incorporated in the standard from day one, without the need for such a huge architecture change as IMS.
There are two main approaches to supplying the ‘missing link’ which would allow non-IMS operators to hand off voice to 2G and 3G networks. One is VoLTE (voice over LTE) from Nokia Siemens, which relies on NSN equipment. It provides a hardware and software upgrade to the existing MSC Server mobile softswitch, provided this has also been updated to support VoIP. This then promises carrier grade VoIP and seamless subscriber mobility between LTE and older networks. Michael Clever, head of next generation voice and multimedia, said customers can later ‘proceed to a full IMS solution to provide converged multimedia services, in addition to VoLTE.’
The other comes from an industry group spearheaded by T-Mobile and several large vendors, called VoLGA. This uses Kineto Wireless’ implementation of the GAN (Generic Access Network) system, now part of 3GPP standards ‘ though VoLGA is not yet working with the 3GPP. GAN was initially the basis of the Unlicensed Mobile Access (UMA) platform, used mainly to bridge Wi-Fi and cellular networks, and is now being used to hand off between 2G/3G and LTE voice.
Of course, eventually cellcos will achieve carrier grade VoIP, mainly via IMS, that meets their users’ expectations of quality of service and mobile hand-off better than generic internet telephony offerings. But most will be supporting two or three generations of network for at least a decade, and will look to hand off the bulk of voice traffic to the efficient circuit switched systems while reserving LTE for heavy duty data.
This pattern has its own challenges ‘ such as making data services available simultaneously with a phone call ‘ but is more attractive to many operators than investing huge sums in full IMS, when the main reason to do so is to support voice, a service with declining margins. While IMS was once seen as essential to support all-IP data offerings, most new-breed mobile internet behaviors access the public internet with no need for the QoS and other features supported by IMS ‘ USB modems, open access and the rise of app stores that are based on the generic web are all reducing the place of carrier-controlled web experiences and IMS. Although IMS is deployed for some specific apps that do need carrier control, such as fixed/mobile convergence and push to talk, increasingly the mobile network is becoming an open internet pipe ‘ whether the operator likes it or not. This leaves voice as the main app that still requires the migration to IMS ‘ not a good business proposition, and hence the new operator interest in interim solutions like VoLTE and VoLGA and future 3GPP support for circuit switched fallback (which could adopt either of the other approaches).
In these days of constrained operator capex budgets, the option of continuing to support most voice activity over traditional networks rather than IMS will be the most popular. The third choice is to open the network to third party VoIP providers, lacking the QoS and emergency call support of a carrier solution but acceptable (and cheap) for many users. The operator reaction to the launch of Google Voice and the Skype Lite handset client indicated that most cellcos are still deeply threatened by such offerings and will keep them off their networks if possible, but consumer demand and net neutrality legislation will prevent this in many territories. Some, notably 3, are grasping the nettle and making high quality Skype support, with the client integrated with handsets, a competitive differentiator.
And others are using fixed VoIP to accelerate fixed/mobile substitution and at least increase their account control and their share of the customer’s total communications/media spend ‘ even if this is falling. T-Mobile USA has been a prime example of this approach with its Hotspot@Home service, and while both AT&T and Verizon have axed their conventional home VoIP offerings, both have fixed/mobile options leveraging VoIP, provided by their mobile arms as part of a fixed/mobile convergence program.
T-Mobile leveraged UMA technology to provide its VoIP service based on its existing mobile core network, using a simple UMA controller instead of needing new voice switching infrastructure. And the system uses the same SIM-based access and authentication system as mobile phones. It is almost identical to Vonage, but with the additional benefit of hand-off to the mobile network using dual-mode handsets, plus integration with T-Mobile’s Wi-Fi hotspot system. This is a key way that cellcos will look to increase ARPU and fend off the VoIP specialists.
Verizon’s attempt revolves around its Hub, which serves as a home phone and central household information store, and runs off an internet line like Vonage. This is sold by Verizon Wireless, not Verizon’s landline unit, as a means to snap up both the fixed and mobile usage of its customers, and to provide homephones outside Verizon territory. This may eat into Verizon landline revenues within its markets, but that’s better than losing these to Vonage, and at least gives the wireless arm a converged account where customers are not willing to cut the landline altogether.
The carriers need to take VoIP seriously, even if they are not prepared to shell out on true carrier VoIP themselves (though some, especially in the CDMA world and notably Verizon, will pursue this route. VoIP support has been implemented in CDMA EV-DO to a more advanced level than in HSPA). According to a new report by Gartner, more than half of mobile voice will travel over VoIP from end to end, by 2019. Gartner expects 30% of this to be initiated or terminated through third party portals like Google, Facebook or Yahoo, all of these having adopted wireless VoIP as one of their basic features. This trend, already kicked off by Google, could sideline Skype and other specialists, and also eat seriously into cellco’s control of voice activity. They have some time to prepare ‘ the report does not expect the bulk of the shift to end-to-end VoIP to occur before 2014, and much will be in the last two years of the decade under study. “Mass scale adoption of end-to-end mobile VoIP calling will not happen until fourth generation networks are fully implemented in 2017,’ said Gartner research director Tole Hart.
Of course, this is just part of the general trend for portals to go mobile and compete directly with cellcos to become the main provider of services, and the main brand, for cellphone users. Google has led the charge, with mobile versions of its key apps, its own developer platform, the Android software store, and the future mobilizing of its disruptive new Google Voice system.
The real casualties of all these changes in mobile voice may be the VoIP specialists, increasingly squeezed between voice offerings from the portals and cellco activities, whether based on IMS or hand-off to older networks. So Google Voice is a challenge to the mobile carriers, but also to Skype and Vonage. In the short term, Skype is biting back by working more closely with the cellcos themselves, to gain the all-important homescreen presence and optimized cellphone experience, and so increasing its competitiveness with other VoIP platforms. The increase in open access will start to reduce the effectiveness of these single-carrier approaches, but they will persist for as long as the open mobile web continues to disappoint in terms of the user experience, compared to its PC equivalent.
Vonage, once the darling of VoIP, has been far slower than Skype in attacking the mobile world, but this year it plans to introduce a mobile client and application, according to CEO Marc Lefar. The company is facing increasing biteback from landline phone providers’ own VoIP propositions, and is suffering under high costs of customer acquisition and low customer additions. One solution will be to bring out a mobile phone client similar to Skype’s, though this may initially use Wi-Fi rather than the cellular connection, at least (presumably) until Vonage can get itself one or more mobile partners to optimize its offering for 3G, as 3 and Nokia are doing for Skype.
This will be Vonage’s second attempt at mobile VoIP. In 2005 it teamed up with UTStarcom to offer voice over Wi-Fi but a single-mode handset without cellular hand-off was quickly outperformed by dual-mode alternatives. Last summer Vonage said it would source a soft client for cellphones from Counterpath, which should be the basis of a full launch this year. Vonage has learned the lesson of focusing on the software and opening this up to any handset maker, rather than providing dedicated phones ‘ this achieves the initial user base that could then attract a manufacturer like INQ to create a really attractive Vonage-oriented phone in future.