The centre has also begun work on network systems architecture, the management of networks and the problem of paying for services on the Internet.

The creation of the Centre for Communications Systems Research can be attributed to the enthusiasm of a group of Cambridge academics who saw the need for research in areas important for the future of the industry, although at first they lacked a full-time director. This became possible when a distinguished Canadian academic, Professor Stewart Lee, agreed to accept the job.

Professor Lee went to Cambridge because he recognised its pre-eminence in areas of computer science, such as privacy and security, and because he believes it to be among the six or seven academic centres of excellence in electronic communications.

"Given its size, that is always astonishing," says Professor Lee. "Unlike many universities, including Toronto, which has 75,000 students, Cambridge is very small. Its total number of students, including postgraduates, is 15,000.

"We felt we could capitalise on Cambridge's excellence but we had to avoid a problem that can easily arise when a company commissions a university to pursue a particular line of research," he said. "Too often the company fails to articulate the problem effectively. Too often the university people, because of misunderstanding or ambiguity, arrive at an answer to the wrong problem. They solve a problem related to the question but fail to answer the question itself.

"We are quite determined not to get into that trap. We are establishing teams of academics and people from industry, whom we call industrial research fellows, who together attack a given problem. The industrial research fellows come with a particular object in mind and are full-time members of a team that will include a Cambridge post-doctoral research fellow and some of our academic members. We have already entered into agreements with Hitachi; Alcatel; the British company, Micromuse, and the US Naval Research Laboratories."

Professor Lee and his Cambridge colleagues are striving to achieve a balancing act between pure research and object-focused research. By attracting industrial funds, which will benefit industry and the university, the centre will be able to push on with other research that in most cases will be relevant to the problems of industry and society.

"It is all too easy for academics to shin up the nearest available ivory tower and disappear from sight," he added. "But that is not what our objective is. Our objective is to keep our feet on the ground, to try to be relevant to the problems that industry is having."

There are many challenges, says Professor Lee. One is to improve the network technology known as Asynchronous Transfer Mode (ATM). Originating in ideas formed about 20 years ago, ATM has evolved into a realistic international standard that enables manufacturers to make their equipment compatible. But there is consensus that ATM could be improved. The Cambridge centre is therefore setting its sights on the future and will be working on ideas for a workable and satisfactory network technology 15 years hence.

Another problem facing the industry is how to charge for services on the Internet. How does a service provider pay for the communications services that it uses? What method can be devised to ensure that charges are linked to usage? The answer, says Professor Lee, is to recruit applied economists.

"The involvement of applied economics is one of our great strengths. In a centre of this kind one would expect to see engineers and mathematicians. But to many people the inclusion of applied economists would be a surprise. As in many areas of science, a multidisciplinary approach is becoming essential."

Professor Lee believes that in some ways we are close to becoming an Orwellian world. It is possible, he says, for unscrupulous people to gain access to the health files of their fellow citizens. "Sociologically this is very serious. We need to take measures to ensure that the illegal gathering of information is not used in areas such as insurance and employment."

One of the centre's projects will be concerned with digital watermarking. Already there are methods being developed to incorporate digital watermarking for the protection of network images that have artistic or intellectual value. All the techniques under consideration involve modification of the picture in a subtle way that the eye cannot detect.

A huge challenge that is just over the horizon is the problem of scale inherent in installing communications systems in the home. Simply doubling a given number of connections is likely to increase some related problems by a factor of four.