Section: Partnerships and Cooperations
International Initiatives
Inria Associate Teams
MOCQUASIN
Title: Monte Carlo and Quasi-Monte Carlo for rare event simulation
International Partner (Institution - Laboratory - Researcher):
See also: http://www.irisa.fr/dionysos/pages_perso/tuffin/MOCQUASIN/
The goal of MOCQUASIN is to design efficient Monte Carlo and quasi-Monte Carlo simulation methods and to apply them to models in telecommunications. Simulation is indeed often the only method to analyse complex and/or large systems, but also suffers from inefficiency. Two specific situations on which we will focus are rare events, and revenue management. In the two cases, we want to deal with dependent individual events or decisions, a realistic situation requiring adapted solution techniques. The inefficiency of the standard simulation is a known issue to compute the probability of rare event since getting it only once requires in average a long simulation time, but most of the literature has up to now assumed independence in the models. The other framework, revenue management in telecommunications, is the situation of providers trying to define valid offers and capacity investments in front of complex demand models. Here too, a change in the decision of an actor has an impact on the others that has to be taken into account.
Inria International Partners
Our other main international partners are:
José Blanchet (from Columbia University) and Peter Glynn (from Stanford University), on rare event simulation
Peter Reichl (from FTW, Vienna, Austria), on pricing and security issues
Héctor Cancela and Franco Robledo (from Uniov. of the Republic, Montevideo, Uruguay), on simulation issues
Alan Krinik (from CalPoly, California, USA), on transient analysis of Markovian queues
Reinaldo Vallejo (from UFSM, Valparaíso, Chile), on solving techniques for Markov models
CNRS/NFSC IRON
Title: Ensuring Interoperability of new generation networks (IRON)
Abstract: Future networks will continue to be heterogeneous. The risk of non-interoperability will increase. This may lead to unavailability of some critical network services, for instance in emergency management, etc. It is important to guarantee that network components will interoperate. One important way among others is to provide efficient testing methodology that help in guaranteeing interoperability of the underlying protocols. The classical testing approach of a single testing system dealing with all tested components and the test execution is no more applicable. To be more confident in the real interoperability of these components, testing has to be done in a close to real operational environment that may be unreliable. Thus, this project aims at providing interoperability testing solutions for distributed communicating systems in unreliable environments.