Today's separation of fixed and mobile networks is a "glaring weakness" in the way enterprise communications systems are set up. As Jeremiah Caron, research director for enterprise telephony at market watchers Current Analysis explains: "The main problem is that the way mobile communications are managed within any enterprise is horrendous. It's a huge financial black hole and nobody really likes the way it's set up. Gaining control over those costs and making them predictable, even if they don't lower them dramatically, is a main aim for any IT department. But it's a very complex process."

Control can only be achieved, say industry watchers, by converging communications media onto one network, giving users one number and perhaps only one handset. That handset must then be capable of switching seamlessly from mobile networks outside the office to wireless LAN inside the office.

The idea of converged networks is not new, says Michael Knott, a managing partner at consultants Accenture, but it is regaining interest as proponents switch "from a technology pitch to a user-centric platform for delivering services that people want."

Convergence presents opportunities for fixed operators to become a ‘one-stop shop' and counter the effect of fixed-mobile substitution says Knott; for mobile operators to diversify into new products and with new partners; and for product and equipment vendors to supply new devices and infrastructure to exploit the full potential of convergence of communications.

But it also presents challenges. These vendors will need to change their business models and balance cooperation on standards with competition for each others' share of service revenues.

Lucent Technologies, which develops communications technology for operators, envisages a converged world based entirely on IP (Internet protocol), backed up by the IP multimedia subsystem (IMS), which allows operators to offer new rich-media services and applications and so gain new revenues to replace those lost as wireless voice and data become increasingly commodity services.

David Poticny, European president of Lucent, says that for a converged network to be viable it has to meet certain quality thresholds before the technology will be adopted by enterprises: users are activated on the service in less than 10 minutes; the device can be used within 30 seconds of being picked up; if a data connection is lost it should automatically pick up where it left off when the signal returns; and the device should always be connected.

"There isn't a single winner technology, it will be a combination [of devices and networks]," he says, but convergence "will be driven by ease of use and ubiquity."

Although several technology suppliers provide products and services that enable enterprises to converge their networks in an ad-hoc fashion (see box), considerable work is required from telcos and network operators to make a converged infrastructure cheap and ubiquitous. Ease of use is just as important on the network management side as for the end user.

Ad-hoc convergence: what organisations can do today

A survey by analyst group Quocirca and mobile service provider O2 found that many people use their mobile phones while at their desk: 11% with high frequency and 44% with medium frequency. UK telecoms regulator Ofcom supports this, suggesting that 30% of mobile calls are made within reach of a fixed-line phone. Many of those calls will be made to co-workers in the same office.

This indicates the growing attachment mobile users have to their personal handsets. Ofcom's statistics also show that mobile's share of switched voice calls has increased from less than 5% in 1994 to around 30% in 2004. Vodafone hopes to extend this trend still further with a business service aimed to substitute fixed for mobile calls. When the mobile giant moved 3,250 employees into its new global headquarters in Newbury, Berkshire, it scrapped its private branch exchange (PBX) network in favour of using only mobile phones, networked to a virtual PBX. The service even includes a mini intranet of internal contacts, accessed via text message. It now plans to offer a similar service to small businesses and mobile enterprise departments.

While that may suit organisations with a remote and constantly mobile workforce, it cannot bring the benefits of full convergence: integrating voice into desktop applications such as customer relationship management software, simplified conference calls or making communications presence aware.

But even without having fixed and mobile voice calls running across the same network, companies can try to converge what they do have as best they can with another software-based approach. ‘Unified communications' platforms, such as Siemens' HiPath OpenScape, Alcatel's Omnitouch and Microsoft's Live Communications Server, sit on top of today's heterogeneous corporate networks to provide advanced collaboration services.

Martin Northend, head of convergence platforms at Siemens Communications, says OpenScape recaptures the fluidity of communication that has been lost as workers become more distributed, by providing a more granular presence function that shows an individual's availability by media as well as by status. "The concept of OpenScape is that it's a communications environment that accepts the fact that you might have five phone numbers and three emails. And the reality is that this is going to be around for some time yet," he says. "You could have a compound icon for one team, so whenever more than half of the team is available, it instantly establishes a voice or data conference with them. If you can do it over IP it provides you with a graphical interface and, if not, a voice-based interface."

In an attempt to conceal the complexity of the current state of wireless communications, some companies have turned to service aggregators. This software, provided by companies such as O2 and iPass, sits on a laptop and controls access to WiFi hotspots, corporate wireless LANs or 3G networks by choosing the most appropriate connection for the user, without any need for fiddly manual configuration. A 3G data card might be perfect for working from a local customer site, but roaming costs could make using it in a foreign airport much more expensive than the local WiFi network. As well as controlling costs, aggregators also bring security advantages, allowing policies to be enforced before allowing devices to connect to corporate networks.

These ideas are small steps towards the simplicity of services, devices and sessions that users need if communications are to be converged, ensuring that the underlying complexity and heterogeneity is hidden from the user.

Unless businesses opt for a converged network, they have to look at overlaying WiFi, says Pierre Trudeau, founder and CTO of wireless network equipment vendor Colubris. But ‘tunnelling' a secure link to the edge switch from a WiFi access point means that the network cannot see inside the tunnel; security and quality of service (QoS) functions then become difficult to manage. "There are two parallel systems to maintain, meaning one more potential bottleneck or single point of failure. The first goal is to leverage 100% of the wired infrastructure to provide a better experience to the user and make it easier to manage by allowing WiFi to be part of the corporate backbone," he says.

Sharing the spoils

Outside the corporate campus, carrier-grade wireless IP voice networks are required to ensure QoS and manage traffic. "Underlying all this somewhere is a network with an operating expenditure," says Dave Keegan, head of ‘office on the move' at mobile operator O2. But so far none of the parties involved have agreed on a business model to share out these costs.

Charging becomes more of an issue as voice over wireless LAN jostles with GSM and perhaps even new wireless access technologies such as WiMax. Industry insiders predict a move towards broadband-style flat rate charging, but this will initially vary depending on the kind of wireless connection being used.

But Peter Linder, CTO of wireline at Ericsson, denies that this will erode operator revenues: "We pay a pound for a thousand litres of tap water at home. But if you add bubbles and a slight taste, you can sell it as mineral water for a pound for half a litre."

He says adding a few ingredients and marketing will prevent flat rates from imposing a ceiling on revenue growth. "As soon as you start exploiting the QoS mechanism it will add or conserve the unique characteristics of each service."

But Current Analysis's Caron is less confident the problem can be so easily solved. "The big issue is the complete lack of conversation between private network infrastructure suppliers and the mobile operators." He says that the ideal of the unified network is being pushed by infrastructure vendors such as Ericsson, Avaya and Siemens but the mobile operators have hardly been involved in working out contracts because they have not yet found a way to sell ‘bottled tap water'. "They still have to figure out a way to keep mobile minutes at an acceptable level to justify the cost of the mobile infrastructure."

Internet telephony platform Skype is one company that is trying a new business model. While basic PC-to-PC calls are free, Skype is gaining currency as an enterprise-ready service by charging for additional functions, such as calling landlines and mobiles, voicemail and conference calling.

These costing concerns have held up development of the end-user technology that supports converged communications. Critics suspect the major mobile device manufacturers are too cosy with operators to hurry out a combined cellular and voice-over-WiFi device, although many such phones are in development. It is also in device manufacturers' interests for users to have one phone for private use and one for business, rather than a single handset that can filter calls using presence functions.

Phones and gateways that allow calls to interface with the rest of the network exist today, made by Hitachi, Spectralink, Motorola and Nokia. But these have not been developed as rapidly as some would have hoped. Mobile companies can point to UMA (unlicensed mobile access), a standard promising to improve call hand-off between GSM/GPRS and WiFi, as proof that they are working towards convergence. But as the industry moves towards IP-based communications in the wireless as well as wireline world, the window of opportunity for this initiative is closing rapidly.

"UMA is just about putting GSM traffic over the Internet," says Colubris's Trudeau. "It is a consumer type service and it doesn't leverage WiFi to the full extent. Enterprises won't like it because it doesn't leverage their IP PBX." By attempting to converge wireless voice only, the mobile operators are pushing in the wrong direction.

As such it has been Research in Motion which, having already started a fire under mobile email with its BlackBerry device, could yet do the same for voice over IP over a wireless LAN, thanks to a partnership with switch maker 3Com. As businesses need a RIM server to use BlackBerrys, the company is better placed to provide the integration with enterprise networks that aids convergence. Nevertheless, indications are that mainstream devices will not be able to provide seamless mobility until at least the end of 2006.

Head-on collision

In the US, it is mobile virtual network operators (MVNOs) that are pushing convergence as a means to increase their market share. But in Europe, the dominance of the old ‘dinosaur' telcos has prevented third parties from edging their way into the market. BT's re-entry to the mobile market as a MVNO using the Vodafone network could break that trend. But its much-hyped ‘Bluephone' fixed-mobile offering, expected by June 2005, uses Bluetooth in its first incarnation – hardly the most seamless convergence tool. A WiFi version will follow, but Bluephone remains targeted at small business and home users rather than as a full enterprise offering that ties into the PBX.

At the heart of all this is the issue of who will "own" converged communications, when it arrives. "There is definitely a lot of positioning and jockeying for brand ownership," says Rob Bamforth, an analyst with market watchers Quocirca. Large telcos such as BT, device manufacturers such as Nokia and software companies such as Microsoft would all like to be the brand most closely associated with the communications tool, he says, "but there is only so much space in the palm of the hand".

Eliminating the line between private networks and those of fixed and mobile operators is crucial to convergence but each wishes to draw their boundary deeper into the others' territories. "The word convergence is not always appropriate," says Bamforth. "It implies something neat. More appropriate would be collision: you get damage. In the fixed-mobile convergence space, people are generally talking about substitution one way or the other."

In the near term, he says, things are likely to get worse before they get better. "We're going through a stage where the potential of VoIP for innovation means that we will see more alternatives," says Bamforth. "You can call it complexity and confusion, or choice, depending on your viewpoint."

Experts at Lucent's Bell Labs R&D unit say that wireless today just cannot compete with fixed line networks, but point to the introduction of high-speed downlink packet access (HSDPA) – the high-speed future incarnation of 3G mobile networks – as the starting point for true fixed-mobile convergence at an operator level. Along with IMS, mobile operator O2 is trialling HSPDA on the Isle of Mann, and it – along with other operators – hopes to offer a commercial service in 2006. While GPRS and 3G are too slow for IP voice calls, putting VoIP over HSDPA, with its download speed of up to 14Mbps, works better than today's GSM wireless voice networks. Only then will that wireless VoIP be able to seamlessly integrate with fixed-line IP telephony.