Sometimes revolutions are not immediately apparent, even to those caught up in them. But those people who attended the 3GSM mobile communications event this Spring can have been in no doubt that something very big is happening in the world of telecommunications.

 “I’ve only seen anything like this twice before,” one IT industry veteran told some of the hundreds of assembled press as he surveyed the vast crowds, the dawn-till-dusk executive programme, the laundry list of product and service announcements. The first time, he said, was back in the early 1980s in the US, just after the PC had been invented. The second was around 1995, also in the US, when the business world suddenly seized on the potential of the Internet.

In each case, these disruptive innovations not only spawned global industries, but changed all industries. Most analysts, operators and suppliers think similar changes are on the way in telecoms. “Telecoms services are about to be revolutionised; it will be like an earthquake,” says Dr Malik Kamal-Saadi, lead telecoms analyst for Informa Telecoms and Media. The way services are offered today will soon be history.”

The coming sea change in telecoms, however, is not exactly news. Telecoms technologies have been steadily evolving alongside the IT industry for the past half century, and the huge impact of convergence and ‘packetisation’ has been widely foreseen. Indeed, the IP (Internet Protocol) revolution has been widely characterised as at least as much a threat to the telecoms industry as an opportunity.

The real reason for all the excitement, though, lies elsewhere, in a much more technical and arcane area: a major new advance is occurring in the deep guts of telecoms delivery software, a layer known as IMS – the IP multimedia subsystem (see article, What is IMS?).

IMS is seen as the weapon that will enable telecoms to regain charge of their destiny and fight off the new hordes of Internet companies that want not only to take their voice business, but deny them the chance to develop new businesses in multimedia service distribution.

Twin spectres

For the past several years, telecoms companies have been looking at the behaviour and the finances of Internet companies – Google, Amazon, Yahoo, eBay (with Skype), AOL, Apple (with iTunes and iChat), Microsoft and scores of others – with a mixture of envy, admiration and fear.

These companies are seemingly able to drop new services into their offerings at will – whether it is video downloads, voice-over-IP calls, or “show me more articles like this”. They measure their lead times to service launch in months, and they can trial new ideas with relative ease. Their services are delivered for next to nothing to almost every desktop and home PC, and sooner or later will be accessible from most handheld devices, including mobile phones.

Moreover, their systems are cheap. They use industry standard Intel servers for processing, application servers, web services to build applications, and IP communication for everything. Skills are in abundance, products highly competitive.

Given that, how can the telcos – with their expensive, proprietary networks and specialist architectures and skills – possibly change fast enough to compete? Is there anything that can stop them from just becoming providers of ‘fat pipes’, utilities over which others offer more profitable and sexier services?

The answer – if it works – is IMS. In the space of a year and half, IMS has moved from its original status as a tentative solution to efficiency and interoperability problems to being seen as the saviour of the operators’ business models, enabling them to deliver new advanced, interactive IP services rapidly and effectively.

“IMS is being seen as something that must be done, and must be done quickly, if the operators are to compete against companies such as Google and Skype,” says Andrew Wyatt, VP of marketing for Apertio, which recently commissioned a report on the take- up of IMS.

“Almost everyone accepts that IMS will enable the operators to compete. If the carriers don’t do something to capture the end users, they will lose them forever,” says Brian Wood, director of marketing for Continuous Computing, a supplier of telecoms software.

The operators appear to agree, and a huge new market has appeared almost overnight. Depending on whose figures are being used, between 50 and 100 contracts have been signed to implement IMS systems in the past several months, and over 100 trials are underway. Informa, which has recently published a report on IMS, says the operators will spend around $4.5 billion a year on IMS equipment and software between now and 2011. Ancillary software, applications, tools and services are likely to add much more to this.

And, as such, individual implementations will equal or exceed the amounts spent on enterprise resource planning (ERP) software and services in recent years.

IMS stampede

Dozens of major suppliers are competing for parts of the IMS feeding trough, including Lucent, Nokia, IBM, Hewlett-Packard, BEA and Cisco, which only recently reversed an earlier decision to pursue a proprietary path.

“Major decisions are being taken now. Deployments are starting,” says Esa Harju, VP of sales for Nokia, EMEA. According to Harju, most of the early sales are coming from the big, fixed line telecoms companies, not the mobile businesses: “It’s driven by the realisation that they are losing customers. Most of the big European telecoms companies are starting to address the issue”.

One of these is BT, whose 21CN network, while conceived and approved before the IMS standards had stabilised, will use the technology where it can.

The key attraction of IMS is it gives operators the ability to rapidly launch new IP services from a common, architectural platform. It means that the operators can introduce and blend new services at dot-com speeds. “IMS is a service creation platform. It’s part of the strategy,” says Paul Wybrow, CTO and CIO for Vodafone, the world’s largest mobile operator. “At the moment, we have to implement things as stovepipe, standalone applications. IMS is the equivalent of a horizontalised architecture.”

Although Wybrow says “IMS is one the many options on our roadmap”, both Ericsson and Nokia recently announced major contracts with the company.

The focus on speed of delivery is key. Vodafone recently launched a Mobile TV service with satellite broadcaster BSkyB which took just eight months to deliver, from announcement to launch. This was viewed as “break neck” speed by the implementation team, who confess that development was often hectic and difficult.

Rob Levy, the CTO of middleware company BEA, says that such projects can be delivered much faster using application server technology (such as BEA’s WebLogic) and an IMS core. He cites one major project at operator O2 that was rolled out in an impressive 16 weeks – but says it can be faster. “New applications should not exceed that timeframe. An IMS network without the services won’t justify the investment”.

But it is not just speed that is important. Once the initial investment in IMS is made, new applications will not only be much cheaper to roll out, but they will portable and interoperable. And they will largely be built using tools and standards familiar to most enterprise programmers.

Most importantly, IMS is built on the Internet SIP (session initiation protocol), which crucially enables peer-to-peer, edge-of-network signalling. This means that users, from almost any device that is SIP-enabled, can access and control their services. Users can be anywhere, on and off the home network, and have their calls and traffic routed to them.

“A lot of people are confused about why IMS is so important. The reason is that existing applications are specific to the network. But IMS is open,” says Brian Wood of Continuous Computing. Operators will be able sell access to applications and services across any network, just as dot-coms do; and, if they wish, they will be able to sell their applications to other operators.

Blended services

For these reasons, IMS has been described as the “the operating system” for the next-generation Internet. And the main IMS systems are being built not by Google or Microsoft but by the same companies that also control access onto the network.

These companies can offer a bundle of services that includes fixed line, mobile and WiFi access, voice, IPTV and a host of other services (see below box, 'IMS-based services'). “IMS gives operators better control over their networks. They can place a semi-walled garden around their subscribers,” says Informa’s Malik Kamal-Saadi.

This explains the optimism that now pervades through the operators, in spite of the threat to their voice revenues. Unlike most of the big dot-coms, they have reliable and trusted billing systems, a strong brand and subscriber base with non-IT literate consumers, and ownership of the ‘on ramp’ to the Internet – whether it is fixed, cellular or WiFi. It may not all go the plan – and the regulators may play a big role in capping their ambitions – but IMS is likely to tip the playing field back in the telecoms companies’ favour.

Further reading

IMS: An undercover revolution - Editor's letter, April 2006

What is IMS? - April 2006

From vision to reality: the convergence of IT and telecoms - September 2005

Life after POTS: telcos' new services - September 2005

More articles can be found in the Communications Briefing Room