Institut d'Électronique et de Télécommunications de Rennes

| ENT |

Accueil du site > Accueil > Production scientifique > Thèses en cours > Équipe SYSCOM (Dpt SC)

Energy Efficiency Scaling Laws for Wireless Communication Networks

Doctorant : Ahmad MAHBUBUL ALAM.

Directeurs de thèse : Jean-Yves BAUDAIS, Philippe MARY.

Thèse débutée le : 01/12/2013.

Recently the “green communications” paradigm has raised in reference to sustainable development for the information and communication technologies (ICT). The wireless world research forum (WWRF) has fixed a 50% decrease of CO2 related to ICT as an objective up to 2020. In order to reach this purpose, the wireless network design criteria need to be reconsidered while keeping an important quality of service (QoS) which is only increasing with time.

Open wireless local area network is the application which is concerned by this project ant the work of the candidate will be included in this framework. In this project, we consider a dense deployment of access points which are directly connected to the internet, avoiding the backhauling from an operator which is very greedy in energy consumption. Previous works have shown that large networks have a low increase of the transport capacity w.r.t. the rise of the number of nodes and the transport capacity per node even tends toward zero. This fundamental limit has been established without any considerations about protocol implementation nor a particular PHY layer. Moreover, this result depends on network topology. Indeed, the authors in have shown that the total capacity could even increase linearly with the number of nodes by using a cooperative architecture. Subsequent works, as in, have focused on the energetic efficiency of ad-hoc networks. More precisely, the authors in have shown that the bits-per-joule capacity can growth more faster than the transport capacity but without taking any into account the unreliable data links. The aim of this work is to study the energy-efficiency metric when the number of access points as well as the number of users vary and also under variable traffic demands.

The PhD program will be separated in three parts. The candidate will start by a rigorous state of the art by studying in particular the previous articles which are very technical. The second phase will consist to study theoretically and by simulation the energy-efficiency in the particular studied network architecture by considering variable number of access points as well as the number of users. The last part aims to study the energy efficiency metric on a network simulator.

Key words : Energy efficiency, cooperative communications, cross-layer, energy-delay tradeoff, scaling laws.

Page du doctorant

Départements et équipes de recherche