A few months ago I wrote about DRITTE and the interesting research work by Umar Saif and team at LUMS. One of their projects is Peer-to-Peer Dialup Networking which aims to mitigate the digital divide by creating efficiencies from dialup Internet. Also called poor man’s broadband, the concept is as illustrated below.
This work, funded by Microsoft Research’s Digital Inclusion Grant, was featured in New Scientist. Note that lack of cheap Ã¢â‚¬Ëœlocal bandwidthÃ¢â‚¬â„¢ in Pakistan is something which has been discussed actively on Pakistani blogs and forums (for example, here and here).
Here is the abstract of the paper:
In this paper we present a peer-to-peer dialup architecture for accelerated “Internet access” in the developing world. Our proposed architecture provides a mechanism for multiplexing the scarce and expensive international Internet bandwidth over higher bandwidth p2p dialup connections within a developing country. Our system combines a number of architectural components, such as incentive-driven p2p data transfer, intelligent connection interleaving and content-prefetching. This paper presents a detailed design, implementation and evaluation of our dialup p2p data transfer architecture inspired by Bittorrent.
For more information see this review at SIGCOMM site. You can also download the paper from there. The authors include Umar Saif, Ahsan Latif Chudhary, Shakeel Butt, and Nabeel Farooq Butt. Read on for this research’sÃ‚Â story in New Scientist.
Here’s an excerpt from New Scientist (subscription needed) article:
IT’S not often that you get to go faster by avoiding the superhighway, but soon students in Pakistan will be able to download big files faster by avoiding the internet.
Instead of using expensive broadband or slow, unreliable dial-up connections, students at Lahore University of Management Sciences (LUMS) will try out a new system, dubbed “poor man’s broadband” (PMB). It allows computers to link to each other directly for faster downloads, and it works as long as at least one computer running the trial software has already downloaded the desired file from the internet. The system should also reduce the university’s risk of overloading the bandwidth supplied by its internet service providers (ISPs).
PMB is a mixture of peer-to-peer (P2P) software – touted as the internet’s future (New Scientist, 13 October, p 28) – and pre-internet techniques, whereby users dialed other computers directly to exchange files. It is based on a peer-to-peer software, called BitTorrent,which allows computers to talk directly to each other and swap chunks of files. The snag is that BitTorrent requires that all computers be connected to the Internet to swap files – a luxury in Pakistan. So LUMS Computer Scientist Umar Saif tweaked BitTorrent to create the new system. Saif’s version allows computers to “gossip” about which computers had what files. Gossiping happens every time computers connect to make a transfer, ensuring the entire network is kept updated.
PMB users still surf the web as usual, but when they try to download alarge file, the software checks its gossip logs to see if it can call another computer directly for a faster download. “Trials so far show this can be more than five times faster than the Internet [alone]”, says Saif.
The system will mainly help in the download of software patches and free educational materials, like MIT’s Open Course Ware, says Saif, because they are large files likely to be found on local computers.
After the university trails, Saif intends to try a city-scale test and hopes to interest other countries with poor Internet infrastructure.