The right connections

Jon Crowcroft and his Communications Innovation Institute team are out to identify the best ways to apply new technology to ensure the industry is not caught by another 3G-style surprise.

Nobody would claim that Vodafone is short of a bob or two. But ever since the company and its fellow network providers paid billions of pounds for 3G spectrum licences, only to find the technology was not ready and people were not desperate to watch football highlights on a tiny mobile phone screen anyway, the industry has struggled to regain its former dash.

‘People have noticed a little bit of stagnation in communications at the moment,’ said Prof Jon Crowcroft, Marconi professor of communications systems at Cambridge University’s Computer Lab. ‘A lot of exciting stuff is going on, but the underlying business is not in brilliant shape and there is a bit of uncertainty.’

Crowcroft is UK head of the Communications Innovation Institute, which was established earlier this year as part of the Cambridge-MIT Institute to investigate emerging communications technologies, and consider the business, economic, regulatory and public policy issues likely to have an impact on their introduction.

As such the institute, involving Cambridge and Oxford universities, MIT, University College London, and BT – which provided £1m of its £3m funding – will be attempting to ensure that the industry is not caught by surprise again.

‘In 3G, we might have had a model that told businesses they did not need to buy that piece of spectrum,’ said Crowcroft. ‘Or told governments that while it is nice to make all that money, perhaps in the long run it would have been better if the spectrum was cheaper, the companies rolled something out more quickly and the government generated tax revenue from that.

‘But actually, I don’t think 3G has a killer application yet, and it won’t have for another couple of years. So the government probably made as much money as it could, because nobody was going to use these things yet, and they managed to fool Vodafone and the other 3G companies into effectively paying early tax. Luckily none of them went broke in the process.’

The researchers hope much of their work will be led by the industry. To this end, they are creating a number of working groups, each designed to investigate a particular technology or area of technology, in which participating companies can sign up to take part as members of the industry forum.

The first of these working groups will investigate broadband technology, and the factors affecting its deployment. ‘One of the problems with broadband is that it could be used very nicely by people who are going to be the last to actually get it,’ said Crowcroft.

‘In very rural areas of the Highlands and Islands of Scotland, or Wales, or the east of England, there are long distances between towns and villages and phone companies are not going to upgrade their copper fibre without a strong argument.’

So the researchers will try to determine the best way to ensure that people in rural areas get broadband internet access. In European countries, particularly the UK, upgrading telephone lines may not be the most cost-effective way of achieving this. Instead, wireless broadband could provide entire villages with high-speed internet access for a relatively low cost, Crowcroft said. ‘There are several technologies that create very fast wireless links along fixed paths. You have a box that costs $50 (£28), one side of which has a fixed wireless link with speeds of hundreds of Mb/s, up to maybe 10km, and on the other side of the box is just wireless ethernet.’

These devices could be placed in a public building such as the village hall, with an antenna on top, and villagers could all be supplied with broadband internet, giving them access to government services and the potential for teleworking, without the cost of upgrading the copper wire or installing optical fibre. ‘It’s in a nice, safe part of the spectrum so there are none of the aerial concerns you get with GSM,’ said Crowcroft.

‘The power is considerably lower in any case, and it is directed out of the village to the nearest city. For European countries – particularly the UK, which has a very dense population – this will connect the last 35 per cent of the country very easily.’

The researchers will also consider how regulation could influence these decisions – for example, Ofcom could insist that companies introducing broadband in cities must also provide a number of rural households with the service.

Another group likely to be of particular importance to the UK will investigate optical communications, an area where the country already has a lead. The optical laboratories at Cambridge, Essex and Aston universities, as well as UCL, are among the best in the world, while fibre optics was invented at Nortel Laboratories in Harlow, Essex.

Optics is particularly interesting as it has over-shot its requirements by a long way. The trans-Atlantic and trans-Pacific undersea fibres are only being used at one to two per cent of their capacity, according to Crowcroft. ‘People in laboratories can get 1,000 wavelengths over 10,000km of fibre, and at the moment that is more than 1,000 times what anyone needs – it’s completely mad,’ he said.Although this sort of capacity may not be needed yet, if deployed it could lead to considerable innovation and this may not be as expensive as people might assume.

‘We have calculations to suggest it would cost around £100 per home, as a one-time cost, to put optical fibre into every home in the UK and Europe, or the US, assuming every home took part. That would cover the cost of setting up the factory to do it, and nobody would ever need worry about bandwidth again. It would just not be an issue.’

However, by providing every household with fibre in one go, optical network suppliers such as Nortel, Marconi, Pilkington and Pirelli would effectively do themselves out of a job and would be left with maintenance work only. So the working group will consider the impact of such a move on optical specialists, including other areas where the technology could be used.

Individual applications will also be investigated to consider which are likely to be popular among users, and the effects they will have on networks. ‘In the past there have been some surprises, the classic example being the success of text messaging,’ said Crowcroft.

‘E-mail is far superior in its ability to send arbitrarily long messages, and a proper keyboard is used – but it turns out that text messaging is still useful.’

The researchers looking into individual applications will establish a dialogue with the companies developing them. These include Urban Tapestries, a small research company investigating a virtual graffiti system, which allows people to walk past a location with their mobile phone or PDA and view a map of the area annotated with remarks other people have made about the place. The company has connected up an area of London with the wireless internet system, and is investigating other applications for it, including neighbourhood watch, said Crowcroft.

‘It requires really snappy performance, because people would expect it to be like graffiti; you could stand there graffitiing an object on your PDA and the person standing next to you would be able to see it, and that’s quite complicated to do,’ he said.

Finding the right applications for a technology is obviously crucial to its success, as 3G has so far demonstrated. Crowcroft believes 3G’s ‘killer app’ will eventually arrive in the form of in-car telematics.

But in the meantime the team is already considering the likely possibilities for 4G wireless networks. ‘We’re really interested in something called mesh wireless networks, which are provided by the community without a company being involved,’ Crowcroft said. ‘So with 4G it is possible that people may just start to connect up to free wireless networks, and the next generations of things like Wi-Fi would include mesh network operation.’

If the government was to persuade Vodafone and its counterparts to pay billions for a network like that, it would be some trick.