2 April 2008--The same kind of peer-to-peer file sharing that made Napster famous--and infamous--is being used in a new research project in Europe that aims to pipe TV programs over the Internet. As part of the P2P-Next project, engineers from several European universities, research institutes, broadcast networks, and manufacturers have agreed to pool their expertise to develop a file-sharing system, based on free open-source software, that could someday allow users connected to the Internet to deliver videos from anywhere to anywhere--and to any number of people throughout the world.
The four-year project, which has attracted more than 20 member organizations, including the British Broadcasting Corp. (BBC), Delft University of Technology, in the Netherlands, and STMicroelectronics, will receive 19 million (US $29 million) from the European Union under its Seventh Framework Programme, with another 5 million to come from the project partners. The goal is to develop not only an entirely open P2P platform for delivering video on demand and live webcast streaming services but one that is also legal, secure, and reliable, according to Johan Pouwelse, a professor at Delft University and scientific director of the P2P-Next project.
The project reflects a growing European interest in Internet-based television, including pioneering work by the state-owned Norwegian Broadcasting Corporation, which has launched a hugely successful TV series delivered via P2P.
Internet video companies like YouTube could someday benefit from the new technology, Pouwelse says: ”Instead of having every bit come from their own central servers, which is costly, they could use P2P to reduce their bandwidth costs.”
Unlike broadcasters, which beam shows from radio masts to home antennas, or cable-TV networks, which send content down a coaxial cable to set-top boxes in a similar broadcast fashion, Internet-based TV providers like YouTube require users to connect to central content servers to fetch programs. Replacing broadcast and cable-based TV service with the Internet would require many more servers, not to mention the strain it would put on content suppliers to provide sufficient bandwidth to transmit the content. P2P technology, according to Pouwelse, tackles this problem by sharing storage and transmission tasks with all enabled users.
The initiative, however, competes against Joost, a commercial Internet TV start-up that largely uses proprietary P2P technology developed by the same two Scandinavian entrepreneurs who launched the Kazaa music file-sharing exchange and the Skype voice over Internet Protocol (VoIP) service. There are also numerous commercial Internet Protocol TV (IPTV) offerings now available from European telecommunications firms, mostly based on technology from Microsoft. Both services have been off to a bumpy start.
Joost, which introduced commercial service last year, has suffered some technical glitches, resulting in frequent downtimes, particularly in March. Users also complain of excessive advertising, which many view as disruptive. And rumors are afloat that the venture could be on its last legs. IPTV has also proven a challenge for many telcos in Europe and beyond. Initial hiccups in deploying Microsoft technology forced a few operators in Europe, such as Swisscom, to delay service; others, including Deutsche Telekom, have yet to find the right business model.
”IPTV is a telco approach with dedicated hardware, a closed business environment, and walled gardens,” Pouwelse says. ”And although Joost uses some open-source for minute components, it's largely proprietary technology. P2P-Next is entirely open to all who want to use it. The system offers more choice and a nearly cost-free way for broadcasters to distribute content.”
Delft University, for instance, is contributing its Tribler technology as a core component of the planned P2P-Next system, according to Pouwelse. Tribler, which stems from the word ”tribe” and refers to its usage of social networks, is a client application based on an open-source implementation of the BitTorrent communications protocol.
BitTorrent, widely used today for downloading TV shows from the Internet, is designed to distribute large amounts of data without the original distributor having to pick up the entire tab for hardware, hosting, and bandwidth costs. Through the protocol, each recipient delivers pieces of data to other recipients, thereby reducing the cost and capacity burdens on any one individual.
Currently, BitTorrent traffic accounts for around 49 percent of traffic on the Internet backbone, of which nearly 50 percent is TV programming, according to Ipoque, a German company that specializes in monitoring Internet traffic.
For years, P2P has been a key technology for content pirates, offering an efficient way for them to share files. Hollywood hated it--until last year when BitTorrent's cofounders decided to go commercial. In a move to win over the studios, as well as publishers of videos, games, and software, cofounders Ashwin Navin and Bram Cohen, the inventor of the technology, added digital-rights management technology to protect content and closed the door to open-source development. Fox, MTV, Paramount, and Warner Brothers have since become supporters of BitTorrent's new commercial service.
Pouwelse believes that the move by BitTorrent's founders to sever ties with the open-source community will, in the long run, undermine further development of the technology, and that licensing fees will deter others from using the commercial application. Fortunately, some components of BitTorrent remain open to implement, he adds, and ventures including P2P-Next are using these to build new systems.
Another advantage of P2P-Next over Joost is its ”zero use” of servers, according to Pouwelse. The system will allow any type of Internet-connected device to participate, he says, adding that the venture will begin with PCs and expand later to other devices. ”By distributing all functionality, we are aiming for unbounded scalability,” he says.
What worked for one hugely successful P2P startup may not work for all, however. With its largely proprietary and somewhat centralized approach, Skype is arguably the most successful P2P VoIP product in the world. The venture found a niche and successfully exploited it. Its business case is now under attack by telephone companies rolling out national and international flat-rate fees. The verdict on Joost is still out.
It's still too early to assess the chances of success for the P2P-Next initiative. Numerous European collaborative research projects have failed or underachieved because of rigid bureaucracy, cross-border rivalries, intercultural differences, or varying opinions on direction. Pouwelse is also honest enough to admit to the various problems inherent to P2P. ”The challenge of P2P is to turn something that can be unreliable and potentially malicious into something that is reliable and trustworthy and works,” he says.
Nevertheless, Pouwelse believes that the initiative's approach--”open source, open papers, and open comments”--could provide a big boost to the project in particular and to the use of P2P technology in general to deliver next-generation Internet TV services.
Others are equally optimistic. ”It's a test bed for new ideas, allowing us to collaborate with colleagues across Europe and to hone and develop technology that could help shape TV of tomorrow,” writes George Wright, executive producer of the Rapid Development Unit within BBC's Future Media & Technology group, in a blog on the BBC Web site.
Pouwelse puts it another way: ”This is really about who will define and deliver the TV standard of the future.”
About the Author
JOHN BLAU writes about technology from Düsseldorf, Germany. For IEEE Spectrum, he explained German resistance to carbon caps on European cars and for IEEE Spectrum Online he described a low-power processor for a disposable wireless vital-signs monitor.