A large part of the results reported in this thesis is based on an observation which has never been made for wireless communications and power control in particular: transmit power levels and more generally transmit covariance matrices can be exploited to embed information such as coordination information and available interference-dependent feedback samples can be exploited as a communication channel. First, we show that the famous iterative water-filling algorithm does not exploit the available information sufficiently well in terms of sum-utility. Indeed, we show that global channel state information can be acquired from the sole knowledge of an SINR-type feedback. A natural question then arises. Is it possible to design a distributed power control algorithm which exploits as well as possible the available information? To answer this question, we derive the characterization of the utility region for the considered problem and show how to exploit this characterization not only to measure globally efficiency but also to obtain globally efficient one-shot power control functions. Motivated by the success of our approach for single-band and multi-band interference networks, we asked ourselves whether it could be exploited for MIMO networks. We have identified at least one very relevant scenario. Indeed, we show that opportunistic interference alignment can be implemented by only assuming interference-plus-noise covariance feedback at the secondary transmitter. Then, in the last chapter, we generalize the problem of quantization, the motivation for this being given by some observations made in the previous chapters. First, we assume that the quantizer and de-quantizer are designed to maximize a general utility function instead of the conventional distortion function. Second, we assume that the quantizer and de-quantizer may have different utility functions. This raises non-trivial technical problems, our claim is to make a very first step in solving them.

Mots-clés : Théorie des Jeux, communication sans fil, allocation des ressources

Composition du jury proposé

M. Samson LASAULCE L2S Directeur de thèse

M. David GESBERT EURECOM Rapporteur

M. Eduard JORSWIECK TU Dresden Rapporteur

M. Mérouane DEBBAH Huawei Examinateur

M. Yezekael HAYEL University of Avignon Examinateur

M. Luca SANGUINETTI University of Pisa Examinateur

Mme Veronica BELMEGA University of Cergy-Pontoise Examinateur

Mme Mylène PISCHELLA CNAM Examinateur