The research team has stated that they have found a way to get around these limitations by using a technique called "Similarity-Enhanced Transfer (SET)". They plan to present the technique tomorrow at the 4th Symposium on Networked Systems Design and Implementation.
SET in itself uses a lot of different techniques to help boost speeds. Many shared files contain peices of identical data. Examples of such pieces are music files that only differ in tags, movies dubbed in different languages and updated versions of applications and software.
SET divides large files into smaller segments first then uses a search method called "handprinting" to find similar files. If it finds any individual chunks that are identical then it adds those to the user's search or current download.
A result of using SET will be an expansion of available sources for any given file. Showing off SET in practice, the software seemed to work well. Using a P2P network such as Limewire, their rate of download for an MP3 track accelerated 70 percent. The results were almost as good when they tried downloading a movie trailer.
The research team hopes to see the software code used in p2p clients soon. "This is a technique that I would like people to steal," said David Andersen of Carnegie Mellon, "Developers should just take the idea and use it in their own systems."
Source:
Arstechnica