Sites Inria

Version française


Françoise Breton - 14/02/2011

IEEE Fellow: a pioneer in game theory applied to networks recognised by his peers

Eitan Altaman © INRIA / Photo Jim Wallace

The Institute of Electrical and Electronics Engineers (IEEE) recently awarded Eitan Altman, senior research scientist on the Maestro project team at Inria Sophia Antipolis - Méditerranée, the title of IEEE Fellow for his contributions in the field of the analysis, optimisation and control of telecommunications networks. The researcher has contributed to establishing and enriching this relatively new field, using tools originally developed for other branches of game theory. Interview.

Why apply game theory to networks?

Eitan Altman:  Game theory can be used in any field that involves questions of strategic competition. It concerns the study of everything from economics to road networks. The emergence of this theory in telecommunications networks (Networking Engineering Games) dates back to the 1990s. A new field of study developed around routing, the control of flows, access and power, and network security. The interest of game theory in the very design of networks became obvious as the number of operators increased and competition became fiercer between service and content providers and network hardware manufacturers. A large number of problems involving game theory arose with the boom in mobile phones. This allows us, in particular, to study and design autonomous networks, such as DTN (where connections are dictated by proximity rather than an operator), while taking into account the problems linked to the limited energy resources of these devices.

Can you give us a few examples of research applications?

Eitan Altman:  We have designed, for example, within the framework of the European Bionet project, routing and energy management algorithms capable of porting applications such as automatic file exchanges over autonomous networks. Another example is the smart telephone network MediaFaun, which I am currently working on with specialists in theatre and IT at the University of Avignon’s new SFR (Federative Research Structure) which focuses on “Sciences and technologies of digital cultures and societies”. We have examined the possibility of supplying multimedia services for major events such as the Avignon Festival or the Fête de la Musique, which gather a large number of participants over a short period of time. These services could involve announcements concerning events in progress or on the verge of starting, with videos that are too large or expensive for mobile networks to carry.

How far has this research progressed and what are its prospects?

Eitan Altman: The field of networking engineering games is new and there are many avenues that remain to be explored. Understanding the dynamics of competition is becoming as important as the notions of balance. An interesting source of inspiration is biology, which also concerns autonomous systems and competition for resources. So-called "evolutionary game theory" seems to offer many possibilities as it applies to dynamic systems that are evolving. In fact, I have contributed to developing some of the theoretical aspects of this approach. Epidemiology is also a source of inspiration in order to understand how a virus spreads or how we can improve a network to make it quicker and more efficient. We now know, for example, how to speed up the spread of information with tools like caches or recommendation graphs.

Another aspect, which has been rarely studied and could contribute a great deal, is cooperative or coalition games. They could identify situations where cooperation can create added-value compared to uncooperative behaviour. This approach is suitable, for example, in dealing with questions raised by Net neutrality, which is widely discussed at the moment and mainly concerns ensuring fair (non discriminatory) treatment of data packets carried by operators (discrimination in terms of speed, cost, protocol, service, packet source or destination, etc.). This can contribute to establishing alternative economic models for Internet access fees and profit-sharing that are capable of stimulating network development, which is a major economic stake at the moment. I am working on these questions in the context of the University of Avignon SFR and the collaborative research initiative MENEUR, financed by Inria.

Some key dates in Eitan Altman's career:

1990:  thesis in Electrical Engineering, Technion, Israel
1990-1992:  post-doctoral research at Inria Sophia Antipolis (Meval project)
1998:  habilitation thesis, Université de Nice-Sophia-Antipolis (subject: Constrained Markov Decision Processes)
1992- ... :  research at Inria and the University of Avignon
2008:  became member of the Alcatel-Lucent Inria joint research laboratory
2010:  awarded the title of IEEE Fellow for his contributions in the field of the analysis, optimisation and control of telecommunications networks.

Keywords: Game theory Networks IEEE Inria Sophia Antipolis - Méditerrannée