After 3G: Voice over IP
By Steve Wallage, Wed Feb 26 10:30:00 GMT 2003

Voice over IP is often presented as a natural evolution in the telecoms market. What'll it take to get there?

Voice over IP (VoIP) is often presented as a natural evolution in the telecoms market. Analyst predictions are that, by 2006, 75% of voice calls will use VoIP. The reasoning can seem compelling. Why have separate voice and data networks? VoIP can be far cheaper and offer far greater functionality both as a 'richer voice' service and also by combining voice and data features.

But there is often confusion about what exactly is referred to by VoIP? Is it voice over the unreliable 'public' Internet or a private, managed IP network? Is it end to end VoIP or simply part of the call carried over an IP network? Is it a proprietary technology or one of the standardized offerings such as SIP, H.323 or Media Gateway Control Protocol (MGCP)?

In the mobile world, there are four main areas of excitement around VoIP.

First, as a way of pitting backhaul traffic over an IP network. This can make eminent sense for mobile operators. Nokia claim that an IP solution can reduce the operational and capital expenditures of an operator by 30% compared to existing switching and transmission solutions. The real figure could be even higher as some operators continue to use expensive and low bandwidth leased lines from fixed operators.

Second, as a way of reducing call costs for customers. TIM Peru signed a VoIP agreement with PRIMUS for its global Internet telephony service. The move was partly defensive as Peru had been the most active PRIMUS VoIP country in Latin America during the first six months of 2002 with traffic growing five-fold in a year. China Mobile offers a VoIP service for its customers. It's a clear case of 'you pays your money and takes your choice' as the operator reminds customers that it cannot guarantee the service quality.

Third, as a way of pushing specific services. A great example is push-to-talk, a hot area at the 3GSM Congress and pushed by companies such as Sonim Technologies and Mobile Tornado. These services depend on the support for SIP, pioneered by such companies as HotSip. This support is still unclear as despite the claimed flexibility advantages of SIP, there is still much operator support for H.323.

The fourth area could be the most revolutionary. This is the idea of using VoIP to integrate cellular and WLAN networks, and provide voice from a WLAN device.

The WLAN Idea

In the same way that companies traditionally reliant on voice lust after data revenues, so (sometimes surprisingly) do data companies often see voice as a great opportunity. A good example is the broadband area. Voice over a WLAN network is a very attractive proposition to many vendors coming from the data side. In fact, some companies have suggested that mobile VoIP could be the opportunity to gain wider VoIP acceptance in the fixed world.

Intel is a good example. As part of its large and growing investments in WLAN, it is funding a VoIP startup, Telesym. Others are working through partnerships. Motorola, Avaya and Proxim have announced a dual WLAN/cellular handset, and are working on seamless handover. Another, older partnership, has been between Ericsson, Agere and Proxim.

Voice over WLAN proponents point to some successful deployments.

In January, Ericsson showcased a healthcare project, codenamed Guardian Angel, using IPv6 and roaming between 2G, 3G and WLANs. This was part of the EU-project IPv6 Wireless Internet Initiative (6WINIT). IPv6 has the huge advantages over the current IPv4, in that it provides a virtually infinite number of addresses allowing huge personalization opportunities. The demonstration carried voice, data and video traffic.

Another working deployment is a school in New Jersey. It deployed voice over WLAN, working with Symbol Technologies, both to save money and improve functionality by also providing student data.

VLI claims that it has more than 200,000 active users of its SIP compliant VoIP software and services for mobile consumer electronics devices such as PDAs.

Getting Back to Reality

But why has VoIP not been a greater success in the fixed world? For many of the same reasons that voice over WLAN will struggle. IDC research found that, despite the attractions of IP PBXs, fewer than 5% of European corporations were even considering their deployment.

There are some clear and reasonable concerns. These include security, management, handover and roaming issues, quality of service (QoS), latency, and the immaturity of some of the standards and technologies. The bottom line is that as voice networks are optimized for voice, and struggle with data, then so the opposite is true.

There is also a major perception problem. Many IT and telecoms managers are very wary of VoIP. They see it as cheap service not meant for enterprise-class voice. Although, users accept some deficiencies in mobile call quality, they will quickly notice if there are increased delays in calls.

A major problem is that the mobile operators are very wary of VoIP. They are worried about traffic and service revenues being taken away. This is why they will, usually successfully, present WLAN as complementary, rather than competitive, to cellular. They can also tariff services to ensure that voice over WLAN becomes less attractive.

A further possible concern is that of regulation. A number of developing countries have been very wary of VoIP. In the US in 2003, the FCC will decide whether or not to regulate Internet telephony. The National Association of Regulatory Utility Commissioners has suggested Internet telephony threatens universal service provision.

Voice over WLAN applications will have lots of trials over the next several years, but large scale deployments will take much longer.

Steve Wallage works and writes for the451. Steve has more than 13 years of experience as a technology analyst specializing in telecommunications.