Jean-Claude Bermond, 2010 EADS prize: architect of networks
Last summer, Jean-Claude Bermond, Director of Research at the French National Centre for Scientific Research, MASCOTTE project-team leader, received the 2010 EADS Enterprise Foundation Prize and the Prize for innovation in distributed algorithmics, awarded by the Sirocco conference. These two prizes recognise his visionary work in applications in telecommunications networks (optical and wireless networks), the design of interconnection networks, configuration theory and the design of efficient protocols for structured communications schemas. Portrait.
"What fascinates me most is the construction of networks, compositions. I love the fun aspect of assembling small structures in larger structures. Bridge, Sudoku, games are part of my life. This is how I work."
Jean-Claude Bermond, barely 24 years old, a young student at a French Ecole normale supérieure and fully qualified to teach mathematics, thus moved towards a career as a researcher at CNRS (French National Centre for Scientific Research), separating himself from his classmates who headed into higher education. That was in 1969. "I first worked in discrete mathematics in Paris, then in Orsay, particularly on graph theory. Then, starting in 1975, I began moving towards computer science . We were quite a minority in the discrete mathematics community to head towards this discipline ", the researcher recalls. "Thus, I became more and more interested in parallel and distributed algorithmics whilst continuing my research on graph theory, then hypergraphs and configurations. Fundamental work that many computer scientists believed was of little interest (unlike programming) because it was too theoretical! " Until the early 1980s, when he began more applied research under a contract with the CNET (French National Telecommunications Centre). "That's where I truly began to use algorithmics on industrial communication problems in distributed memory architectures. "
Leaving and returning... to Inria
In 1987, tired of the administrative red tape associated with his research, he needed to take a step back. He headed to the west coast of Canada, to Vancouver. Upon his return a year later, he approached the I3S laboratory, established in Sophia-Antipolis... and soon became director! "Very quickly, I approached Inria, particularly two directors, Pierre Bernhard and Gilles Kahn, in order to try to pursue a joint policy. At the time, we were already discussing the creation of an ICST campus! " The ties strengthened over the months and led to the emergence of the SLOOP project in 1995, then the MASCOTTE project-team in 2000. "Thanks to my Inria colleagues, I was able to intensify the industrial collaborations with the telecommunications world. " Gilles Kahn enabled him to meet with the Toulouse Alcatel Space team. This partnership led to the design of embedded fault-tolerant networks, a patent filed with Alcatel, and the design of the networks used in ASTRA satellites.
"Industrial partnerships played a key role in the course of my work. While we, the researchers, contribute our know-how and knowledge, the industrial players provide us with interesting problems and a different way to approach research. The blending of these two worlds is very productive." The main argument in favour of this approach? His own research work, which resulted in industrial applications either immediately or within a few years.
Transferring and transmitting
Although it is important to be able to disseminate knowledge in the industrial world, it is equally important to disseminate it to students. "I like teaching. I have supervised more than fifty students in their theses. Their energy and vitality are fundamental in research. Because there are still many issues to resolve. For example, the Internet network and radio networks are commonly used even when not all aspects of their operation are scientifically mastered. Research is almost falling behind industry. " Another avenue to explore: the development of distributed algorithmics making it possible to work with partial, non-comprehensive information (more consistent with real data than the centralised, very theoretical view). Or the design of graphs that change over time, made necessary by the arrival of new telecommunication technologies. Or even the construction of energy-efficient complex networks. "The design of green computing is a true challenge for the coming years. " Jean-Claude Bermond proposes these new challenges to the young scientists who will perform the research of tomorrow. At age 65, this marathon veteran, ultra-trail enthusiast, is handing over and conveying his enthusiasm to the next generation. With this record, he will be a hard act to follow...