1st class CNRS researcher at the computer science laboratory of Grenoble LIG (Laboratoire d'Informatique de Grenoble) , group DRAKKAR.
My general area of interest has been for a long time computer architecture, and more specifically the interface between software and hardware.
From 1982 to 1993, my main scientific interest was in the Prolog programming language. My first project in this area was the design of a Prolog Data-Base machine (OPALE project). Then, I oriented my researches toward the study of the impact of Prolog on the architecture of processors (instruction sets) and memory hierarchies (cache, virtual memory).
From 1982 to 1993, my research has focused on the Prolog programming language. My first research project in this area has been the study of a Prolog database machine (Opale project), in which we developped an algorithm to allow the unification (basic operation in Prolog) of a set of goals (requests), whith a set of rules stored in secondary memory.
Then I directed my work to the study of the impact of prolog on processor architectures (instruction set) and memory hierarchies (caches and virtual memory).Since 1992, the evolution of the research area, and the growing importance of networks (and especialy ATM) led me to focus my interest on this area. My first works consisted in the study of the impact of the ATM on massively parallel computer architectures. As the interest for this kind of machines was decreasing, I switched to the networks of workstations approach, based on message-passing interfaces such as MPI.
The works done in this domain shows interesting perspectives, but communication performances (particularly latencies) was limited by software overheads. I henceforth started the Median projet (Message processor for Efficient DIstributed processing on ATM Networks), which aims at designing a hardware communication interface allowing to improve communication efficiency for distributed or parallel applications running on an ATM network of workstations.
My current scientific interest are now oriented toward the more general domain of computer networks, and more specifically wireless networks. In this domain, I am interested in the methodology of experimental evaluation of the performances of wireless network implementations. In particuler, I developped a method and tools to allow the verification of timings at the MAC level through the observation of the statistic distribution of interarrival delays of packets sent by a continuously transmitting station. This method allows to achieve a precision in the order of the microsecond (and perhaps better) with off-the-shelf hardware and software. It can be used for example, to check the conformance to standard of experimental implementations, or free software implementations of drivers.
In this context, I developped ipmt, (Internet Protocol Measurement Tools), a suite of tools allowing to generate flows of packets in UDP or TCP with given characteristics (packet sizes, delay between transmissions, burst size, data rate...) and to observe continuously received data at destination (receive bit rate, lost packets of jitter in UDP...). Theses tools are regularly used for our experimentation, and by students for practical works.
Im am also interested in protocol for time synchonisation on computer networks, such as NTP or PTP.
Born on 10 april 1951, in Lyon (France).
Maried, three children.
Hobbies : hikking, fishing, photo...