My previous news post on the new age of P2P (Speed Up Your P2P) has been getting some interesting traction from search engines and the like, so in that stead I’ve decided to post a little more on the concept behind Similarity-Enhanced Transfer P2P.

First off, our basis of understanding of what SET hopes to be is based on this comment from CNET —

“Computer scientists at Carnegie Mellon and Purdue say they’ve conquered one of the frustrations of P2P downloads: the sometimes painfully slow speed.”

SET is to be aimed at the sharing of academic papers throughout the community; and will no doubt have other uses given enough time. They have already been quoted saying it would be effective in the transfers of large data files, for example movies and music.

EurekAlert has an article which describes it in more detail, but essentially (to my understanding) the system does the following: locate chunks of data that are similar to the required data for your download. This means that it increases the amount of sources available for a given P2P steam (like in the bittorrent protocol – more sources = more throughput). For example if you were downloading a whole movie, it could download “parts” of it from people sharing the trailer to the movie.

In some cases, SET might speed transfers by just 5 percent; in others, it might make downloads five times faster.

Anderson hopes the technique is used elsewhere, across all relevant P2P networks. Anderson says “developers should just take the idea and use it in their own systems.”

In tests based upon real files downloaded from today’s peer-to-peer networks, SET improved the transfer time of an MP3 music file by 71 percent. A larger 55-megabyte movie trailer went 30 percent faster using the researchers’ techniques to draw from movie trailers that were 47 percent similar. The researchers hope that the efficiency gains from SET will enable the next generation of high-speed online multimedia delivery.

For more information see EurekAlert and CNET.

Popularity: 11% [?]