French version of this page French version of this page
This action aims at studying routing (the strategy for establishing a way
for a given message through the network) and handover (when an end-user
changes its relay station, moving from one covering cell to another) questions
in mobile ad-hoc networks and LEO satellite constellations.
Mobile ad-hoc networks. Given a random field, we consider a relatively dense population of mobile sets. Each set can be considred either as an end-user or as a relay. We aim at connecting these mobiles one to another, or connecting a mobile to a neighboring network by wireless connections and routing among the mobiles. The network structure becomes dynamic and unpredictable, and we try to get the best of it in a real time fashion.
LEO satellite constellations. LEO satellites are orbiting at constant speed and constant altitude fixed between 500km and 2000km. As a consequence, the covering cell of a satellite moves continuously. As a consequence, in order to provide global coverage, we need several LEO satellites describing a set of regularly spaced orbits. The most popular examples of these systems are probably Globalstar, with 48 satellites on 8 orbits, and Iridium, with 6 polar orbits of 11 satellites each. Communications links provided by these constellations are bi-directional, and be of type satellite-phone, satellite-gateway, and eventually among the satellites, which allows routing inside the satellite network, without transiting through terrestrial networks.
In these networks, the restricted geographical covering of each station make locating questions very sensitive to the users' movements, which are hardly predictible. Moreover, in satellite constellations, the covering area of a LEO satellite changes continuously, since the stations themselves are mobile. In any case, it implies that during a communication, an end-user can have to use different stations. Within this context, we study routing and handover questions. We found useful to merge our research effort on those traditionaly different fields.
We stress that nowadays routing is no more based on static features.
Time inhomogeneity is taken into account (varying tarif rates, use of resources
by night, etc.), along with space inhomogeneity (urban and rural zones,
organization of great events). Intersatellite links and inter-mobiles links
make the structure of the network dynamic and complex. Finaly, these networks
could also be employed to cover classical telephone users of isolated rural
zone, which would make the shape of the traffic very unusual.