by Christian Huitema
The main objective of the project is to design, implement, and evaluate mechanisms that will allow the efficient and flexible use of high-speed networks (with a bandwidth greater than 1 Gb/s). To achieve the above objective, we focus our activities in 2 main areas, namely: transmission control protocols for high-speed networks and high performance applications.
We are designing and developing flexible and high performance transmission control protocols which integrate mechanisms for the allocation of resources in the network, for multicast delivery, for flexible error control, etc. These protocols are used in particular in the INRIA Videoconference System (IVS), an application for videoconference over the Internet.
Previous research in the RODEO project led to the development of the MAVROS compiler, and of fast algorithms for the management of data transparency in heterogeneous systems. We are extending this work so as to allow the compiler to handle detailed specifications of the applications. This will lead to further optimizations.
We expect that the two research areas described above will eventually merge within the framework of the recently proposed Application Layer Framing (ALF) architecture. It will then be possible to take into account specifications of an application, of the network ressources required by the application, and to automatically generate tailored communication modules that handle both data transparency, synchronization, and transmission control for this application.
RODEO was a member of the Esprit project OSI-95 along with Bull, Olivetti, Alcatel, Intracom, and the universities of Liege, Lancaster, and Madrid. This project led to the design of a transport protocol for high-speed networks. Work on a high performance protocol architecture is ongoing, within the scope of the HIPPARCH project started in January 1994. Our work with IVS, the videoconference application for the Internet, is done in part within the the European MICE project, and within Telesia (with INRIA Rocquencourt). We are also involved with the European Value project PASSWORD which deals with network security. We are active in various working groups in the IETF (Internet Engineering Task Force), in particular those groups dealing with network protocols and multimedia applications. The MAVROS compiler (developed within the Esprit project THORN) has been licensed to E3X and Bull, and further work is done in collaboration with Siemens.
The HIPPARCH project, A European-Australian Collaboration
The design of flexible and efficient protocols is an important step toward the building of high performance distributed applications over high speed networks. This project proposes a novel architecture for communication protocols. The global objective of the project is to build a development environment which automatically generates the communication module required by a distributed applications using application-specific knowledge to tailor this communication module for improved performance. The architecture will be validated by prototype protocol implementations and test application demonstration. The architectural design is based on the Application Level Framing and Integrated Layer Processing concepts -- techniques for integrating protocol implementation and reducing communication bottlenecks.
The proposed project will address the design of efficient and application oriented communication subsystems over high speed networks. The main objectives can be specified as follows:
The following phases in our methodology can be identified:
The MICE project
The MICE project, coordinated by UCL, addresses these requirements. It aims to provide multi-way integrated conferencing service between a number of European pilot sites, and link them to the appropriate communities in the US. MICE demonstrates a range of possible ways of interworking between the project partners who are using heterogeneous environments, rather than develop new systems or a special infrastructure. The intention is to interconnect up to half a dozen conference rooms, and up to a dozen workstations both by the ISDN and research networks, and to pilot multi-way conferencing services between them. The project has demonstrated multi-way conferencing between 11 sites in 7 countries within 8 months (though only four at a time), and will demonstrate multipoint with up to 4 simultaneous users from 12 sites in 7 countries within 12 months.
The initial partners in the MICE project are: ULB (Belgium), INRIA (France), GMD (Germany), ONERA (France), University of Stuttgart (Germany), NTR (Norway), University of Nottingham (UK), University of Oslo (Norway), SICS (Sweden), UCL (UK). In addition, two National Research Network Operators (the German DFN and the British JNT) are associated with the project without funding. Both they and the CEC are planning to attach their own conferencing systems for interactions with ones provided in MICE. Others sites have asked to join the project, and the consortium would welcome further partners if the project is able to continue beyond the initial 12 month period.
The MICE project relies heavily on developments, expertise and facilities provided in other national and international projects; only the international aspects, the technology integration, and the operational coordination are financed by the CEC. Therefore, the project uses a range of existing conferencing equipment from other projects, which have a major bearing on the choice of subsystems used. With so many countries and public monies were involved, the project has, naturally, chosen to adopt international standards wherever possible.
inria.fr