International Internship Program 2006


(stages d'octobre 2005 à mars 2006) / (Internships from October 2005 to March 2006)


Votre courriel/e-mail of the sender : Mariette.Yvinec@sophia.inria.fr

Nom du projet/Research team name : Geometrica

Unité de recherche/Research Unit : Sophia-Antipolis

Thème INRIA/Research theme : Sym

Nom et prénom du chef de projet/Research team leader name :  Jean-Daniel Boissonnat

Encadrant du stage/Intern tutor : Mariette Yvinec

Titre du sujet du stage/Internship title
Voronoi and mobile phones

Type de stage/Intern level : diplôme d'ingénieur-Engineering school    or      Mastère-Master's thesis 

Durée minimum du stage (en mois)/Internship duration (months) : 6

Ce stage pourrait-il déboucher sur une thèse ou un post-doc ?/Eventual follow-up : oui-yes

Description du sujet du stage/Internship description (une dizaine de lignes/about ten lines) :

Given a set of antenna for mobile phone, we want discretized the space depending on the best antenna visible from the phone. Such a modelization can be done in different way for different purpose, precise wave propagation simulation gives good result to simulate the actual behavior but is not very useful to really design the network. We are interested here in a more abstract modelization using a Voronoi-like diagram. Unfortunately, the anisotropic metric associated to antennas make the direct computation of such a diagram difficult.

We intend to develop a discretization solution, where a mesh of the diagram is constructed. From a course mesh, if an element is evaluated to belong to several Voronoi region, then an oracle will produce points on the boundary of Voronoi regions to add them in the mesh.

The student will develop a software solution based on the library CGAL (www.cgal.org). Relying on already existing meshing solutions. He/she will have to design and develop the oracle. This work will be done in collaboration with France-Telecom research.

Préciser les pré-requis nécessaires pour ce stage/Pre-requisit :
Good knowledge of C++ and the STL
Classical algorithms (e.g. sort)
Knowledge of Computational geometry is a good point but not mandatory


© INRIA - mise à jour le 31/05/2005