Single minded

Engineers are planning a single system to handle all electronic communications at complex sites such as airports, factories, hospitals and universities.

The idea is to cut infrastructure costs by uniting dozens of different wired and wireless networks, and the solution will also support mass monitoring of RFID tags.

The new systems will be intelligent, self-organising, adaptive and able to handle so much data that they will track everything that moves. Data for disparate applications such as speech, CCTV, terminals, tags, vehicles and displays will be hosted on the same network.

'It can be very expensive to install, maintain and operate all the different networks that are used at the moment,' said Prof Jon Crowcroft of Cambridge University, which is leading the project. 'You've got a whole bunch, including 2 and 3G cellular networks and wi-fi.

'The problems caused by the variety of systems can also be compounded by the extent of the coverage. For somewhere like an airport you must have wireless access throughout the buildings, into the public areas and out on to the runways,' he said.

Crowcroft pointed out that each network requires its own infrastructure which is not only robust in itself but also must not interfere with any of the others. The bottom line is that these requirements have their own costs which would be largely eliminated if all networks could be combined.

So the multimillion pound project has won funding from the government's EPSRC. Cambridge has commercial partners in the project, including Motorola, developers Laing O'Rourke, BAA, Boeing and installation planners RED-M.

As well as reducing network costs, the likely solution also has potential to accommodate a growing number of computing and communications tasks. It is focusing on two technologies that have been refined recently — fibre radio and active RFID.

Fibre radio uses optical fibres to carry radio signals from one part of a site to another. A novel piece of equipment has been developed at Cambridge and University College London to sit at each termination of the fibre to receive and transmit the signals. 'This technology is very cute because it is "protocol agnostic". It doesn't matter which radio technology you have,' said Crowcroft, 'It can transceive any signal whether it's a cellphone, a wireless ethernet or something else.'

This means the same box can be installed at every fibre termination around the entire site and the separate technologies that are needed for every different radio protocol will be needed just once each, deep in the machine room at the network's centre. This reduces complexity and could cut costs significantly.

The second advancing technology which the project will exploit is active RFID. 'Radio frequency identity tags are best known for their application as passive devices in retailing,' said Crowcroft. 'That's a fairly simple task because the tagged items are fairly static. But it's more difficult when you are trying to keep track of hundreds of thousands of items of luggage, equipment, vehicles and people — many of which are moving.'

Fortunately, recent work at Swansea and Cambridge universities into active RFID, using powered tags, offers a solution, allowing many tags to be pin-pointed as frequently as twice every second.

These active RFID tags will operate at low data rates, typically 64kbit/s, but an airport can be expected to contain a few million of them. Combine this with the data volumes of all the other networks and the single unified network may have to handle up to 100 Gbit/s in a relatively local access environment.

This is well beyond the capability of any current network, making the fibre radio solution even more attractive.
Crowcroft said the three-year project kicks off officially in October and, unusually, began when it was approached by companies working on the new terminal at Heathrow.

'The engineering companies building Terminal 5 actually approached [us],' he said. 'They said they had a problem and had heard that we were working on something that might match [their needs].

Then the funding opportunity came up. When three things arrive at the same time — our existing research, the commercial need and the funding window — it must be a good thing to go for.'