There's a big gap between ADSL and fibre, and fixed wireless broadband is the technology to fill it, according to the Robert Condon, the chief executive of Libera, the company that just launched its plans for 36Mbit/s wireless broadband for businesses in the UK.

"The time has never been better to launch this," he told Techworld at the VON Europe conference in London. Despite possible confusions with WiMax services, and the fact that hardware for the service is still costly.

Ubiquitous broadband has produced an appetite for high speed data that can't be satisified by 1 or 2 Mbit/s services, he said. Despite the failure of earlier point-to-multipoint services, prticularly in the US, things have changed: "Wireless really does perform well now," he said.

Fixed wireless isn't for yokels
Other companies (including BT) are proposing fixed wireless as a technology to fill the gaps in ADSL coverage, particularly in rural areas. Condon reckons this is missing a much bigger opportunity: for businesses that already have broadband, but want more.

Hence a service that is starting in London, where there are many medium-sized businesses already on broadband. Business broadband services don't offer very reliable services, and the bandwidth is low compared with what users want: "True broadband is considerably more than one or two Mbit/s," he says.

Unfortunately, the only alternative is fibre-based services, also usually from BT, which are hugely expensive, and usually beyond the reach of small businesses - in both senses. "Fibre doesn't usually run where you want it," he says. London has a lot of fibre, but Libera sees a lot of opportunity in the "doughnut" of businesses outside the very centre of London but within the M25.

Over the next year, Libera plans to offer wireless broadband up to 36 Mbit/s within the M25, starting with a trial in Docklands this summer. Beyond that, the company has enough licences in the 28GHz range to cover 50 percent of the businesses in the country, including other urban areas such as Manchester. It can take its coverage to 75 percent of businesses with unlicensed spectrum, according to a FAQ on the Libera site.

"We are the first real alternative to BT's local loop," says Condon, and he intends to make the most of that comparison, by offering voice services as well as data over the service, including a hosted PBX service. He quotes Analysys figures that say there will be a £350 million market for voice on IP in the UK by 2008: the detail of the bar chart makes it clear that the majority of this is in low-cost VoIP aimed at enterprises (see our feature coverage of IP telephony).

What are the problems?
Some people will confuse this with WiMax (see our articles on the WiMax hype), but that is not a problem, says Condon. "We are complementary," he says. "We work on licensed spectrum."

Another problem - addressed at some length in the company's FAQ, is the perception that wireless communications is unreliable and insecure. A few well-publicised users should counter this problem however.

Like any roll-out of new technology, Libera suffers the problem that the technology is a moving target. The customer equipment for the full 36 Mbit/s currently costs around £2000, says Condon, while equipment operating on the unlicensed 5.8GHz spectrum only costs around £300.

However, the backbone equipment is much cheaper and more reliable than it used to be, he says, which will keep the market from the kind of carnage that took place in the US when Metricom and others filed for Chapter 11 bankruptcy protection. "Those companies paid heavily for licences and had a huge commitment in building their network," says Condon. "Just when they needed the money, the market fell off the edge of a cliff and funding dried up."

So will it work?
Condon exudes confidence in the proposition. As a self-styled serial entrepreneur, this is his stock-in-trade, of course, but his track record is good, with senior positions at Cable and Wireless, as well as launching both Atlantic Telecom and Broadnet Europe.

The copper-fibre gap he identifies is a real one, which is inadequately addressed by services such as SDSL, and urban business broadband is a more productive target than rural broadband.

The proof of the concept will be the performance of the trial, and the costing of the eventual service.