En poursuivant votre navigation sur ce site, vous acceptez le dépôt de cookies dans votre navigateur. (En savoir plus)

PhD student in computer science (H/F)

This offer is available in the following languages:
Français - Anglais

Date Limite Candidature : mercredi 30 juin 2021

Assurez-vous que votre profil candidat soit correctement renseigné avant de postuler. Les informations de votre profil complètent celles associées à chaque candidature. Afin d’augmenter votre visibilité sur notre Portail Emploi et ainsi permettre aux recruteurs de consulter votre profil candidat, vous avez la possibilité de déposer votre CV dans notre CVThèque en un clic !

Faites connaître cette offre !

General information

Reference : UMR6074-ANNORG-002
Workplace : RENNES
Date of publication : Wednesday, June 9, 2021
Scientific Responsible name : Anne-Cécile Orgerie
Type of Contract : PhD Student contract / Thesis offer
Contract Period : 36 months
Start date of the thesis : 1 October 2021
Proportion of work : Full time
Remuneration : 2 135,00 € gross monthly

Description of the thesis topic

Modeling and optimizing edge computing infrastructures and their electrical system

Connected objects are progressively invading our daily lives with ever-expanding fields of application : personal health equipment, smart buildings, smart grids, connected vehicles, smart cities, etc. A recent study estimates that there will be 50 billion connected objects by 2025, with significant economic benefits in the health, energy, automotive and construction sectors. All these objects, connected to telecommunication networks (usually the Internet), can interact with other connected objects or with distributed computing infrastructures, such as Clouds for example, to store information, share data or perform calculations.
For low-latency purpose, the Cloud resources are pushed towards the edge of the communication networks, thus being closer to the connected objects than usual Cloud data centers. This emerging Edge computing architecture is characterized by its very large scale, dynamic nature and heterogeneity.
The growth in the number of connected objects and supporting infrastructures poses scientific challenges, notably concerning the management of scaling, the heterogeneity of the communication networks used (Ethernet, WiFi, 3G, etc.), the migration of calculations between objects and supporting infrastructures, and their worrying power consumption. As edge computing constitutes an important shift towards massive decentralization of Internet infrastructures, this change is accompanied by a growth in the number of computing and storage resources and a greater heterogeneity of these resources, which increases the complexity of effectively managing these infrastructures. Moreover, the geographical dispersion of the Edge resources offers locally limited computing capacities and adds a dimension to the already complex problem of the global optimization of the infrastructure's energy consumption. Indeed, this growing and worrying energy usage worsens the environmental impact of Edge infrastructures.
Today's Cloud providers are trying to reduce this impact primarily through two techniques : 1) improving energy efficiency, through improved cooling systems in data centers for instance, and 2) using energy from renewable sources, primarily through photovoltaic panels managed by the Cloud providers themselves. Yet, renewable sources often offer intermittent and variable production on given geographical sites. In this PhD thesis, we aim at modeling and optimizing edge cloud infrastructures and the electrical system powering them through renewable energy sources.
The goal is to accurately model the edge computing infrastructure with its users, their impact on the infrastructure load and the quality of service they individually require on the one hand, and the associated power system on the other, as these two parts are interdependent. Indeed, the workload of the distributed edge computing infrastructure directly influences its power consumption and the power system, with its renewable energy sources and energy storage facilities, influences the load allocation policies in the edge infrastructure, policies that aim to optimize its energy consumption.

Work Context

This PhD will be funded through a CNRS PhD grant and hosted in the Myriads team at IRISA laboratory in Rennes. It will include collaborations with the SATIE laboratory in Rennes. It will be supervised by :
— Anne-Cécile Orgerie, CNRS, IRSIA, anne-cecile.orgerie@irisa.fr
— Anne Blavette, CNRS, SATIE, anne.blavette@ens-rennes.fr

Constraints and risks

The required skills are :
— Master degree in computer science or equivalent
— strong knowledge in distributed systems or telecommunication networks
— knowledge in modeling and simulation, or optimization techniques
— knowledge/interest in renewables and power systems would be appreciated
— strong taste for research
— programming skills
— excellent communication and team skills
— writing and reporting skills

We talk about it on Twitter!