"The distribution of digital media and audiovisual content has recently seen many developments," says Denis Mischler, a researcher at Thomson, which is leading the MediaNet consortium.
The result is a myriad of new online services over shared access and home networks, from interactive TV to multi-channel publishing. But fully deploying all these potential applications creates numerous challenges.
The partners concluded that today's technologies for video coding and streaming are inadequate, notably the popular MPEG2 used in traditional broadcasting. "We need more efficient coding and technologies to handle multiple media formats – including the ability to stream high-definition TV to small devices such as PDAs," says Mischler.
Shortcomings were also found in the areas of content security and copyright protection, seen as vital components in the e-media value chain.
"Future networks will convey a huge mass of content, but only if digital content owners feel their rights are adequately protected," says Mischler.
Home networking also throws up a number of challenges. These include managing bandwidth and QoS (Quality of Service). For example, a network carrying a variety of services – such as films, TV programmes, games and music – will result in a mix of QoS requirements. Yet the end-user's sole concern is to get reliable services without breaks.
Ideally, there should be a smooth exchange of digital content throughout the e-media chain. The project's solution is to improve the interfaces between or within each of segment in this chain.
"At the top level is the reference architecture," says Mischler. "We identified some key interfaces for different e-media stakeholders to work together and provide their services, without handling everything from service provisioning to the end-user," he adds.
"Service providers can deploy services in an open environment. Network access providers can add value and generate extra business. And end-users benefit from this openness – they don't need to buy everything from a single provider."
There will be indirect and direct benefits, believes Mischler. First, by opening the value chain, this solution would offer end-users more choice and at lower cost. It would also let them access more service providers. Second, it would help the audiovisual sector – which today offers no guaranteed QoS – by completing technologies such as content compression. This would, for example, enable quality guarantees in analogue TV broadcasts for use on IP TV.
The partners assessed QoS in the home, where guaranteeing bandwidth is not under the service provider's control. "With service providers, we implemented home-network capability protocols. These allow different devices to access the network, including personal computers and consumer devices," adds Mischler.
Research work looked at areas such as QoS, architecture, compression and telephony. The most notable result came with H.264, a high-compression video standard which can encode video with three times fewer bits than current MPEG2 encoders used in TV broadcasting. MediaNet's H.264 standard has been adopted by the Joint Video Team of the International Multimedia Telecommunications Consortium (IMTC), responsible for the MPEG4 Advanced Video Coding/H.264 specifications. This standard is being rolled out across Europe for digital video broadcasting and broadband Internet video-on-demand services.
MediaNet also contributed to develop audiovisual evolutions of standards for the next generation network derived from the 3rd Generation Partnership Project (3GPP). The mobile telephony services initiative has also approved the inclusion of AVC/H.264 as an optional feature in its latest mobile multimedia telephony services specifications.
To verify whether real applications could work over the project architecture, they demonstrated several 'use cases'. Among them were integration of advance telephony services with TV over IP services, automation of content publishing for many devices –from a phone screen to a wide-screen TV, and content protection for a digital home network.
While there is no guarantee this open architecture will be taken up commercially, some of the partners are already exploiting various project results. In terms of compression, for example, Thomson plans to use H.264 and pursue its development in the IP world. Others may build on the project's prototype security framework for end-to-end content distribution.
Contact:
Denis Mischler
Technology group - Corporate Research Rennes
THOMSON R&D France
F-35576 Cesson-Sévigné
France