A different approach to organizing information on the Internet could lead to a faster and more energy efficient network, according to a research team that included the University of Cambridge Computing Laboratory.
The European Union Pursuit project, which concluded earlier this year, is a proof-of-concept model for an overhaul of the Internet's IP layer, through which isolated networks of computers are internetworked. The claim is that a radical change in the way information is stored online will make it faster and safer to use.
Under the Pursuit scheme, users would be able to obtain information without requiring direct access to the servers where the content is originally stored. This is because copying and caching of data would be a fundamental part of the operation of the data access process. It would be similar to the peer-to-peer networking used by some file-sharing sites but extended across the whole of the Internet.
It would potentially make the Internet faster, more capable of meeting escalating levels of user demand, and virtually immunize information delivery against server crashes. The researchers also argue that by focusing on the data, which would necessarily be marked or fingerprinted, digital content would be more secure.
The Internet currently operates essentially by doing a global search to find a source of information, whereupon the client computer calls up the server computer with packets of information moving between them from end to end, according to Dirk Trossen, a senior researcher at the University of Cambridge Computing Laboratory. This is rather like the set-up and tear-down of a telephone call.
However, users aren't much interested in the storage location or connecting the endpoints. They just want the information, said Trossen in a statement. Technically, online searches would stop looking for URLs (Uniform Resource Locators) and instead look for URIs (Uniform Resource Identifiers).
These would be identifiers that would enable the network to work out what the content is, which would have particular relevance for much downloaded content.
"Widely used content that millions of people want would end up being widely diffused across the network. Everyone who has republished the content could give you some or all of it. So essentially we are taking dedicated servers out of the equation," said Trossen. "The need to do something like this is only going to become more pressing as we record and upload more information."
The Pursuit project ran from September 2010 to February 2013 at a total cost of 5.2 million euros (about $7 million) with about 3.77 million euros (about $5 million) provided by the European Commission.