You are here

Microclouds: An Approach for a Network-Aware Energy-Efficient Decentralized Cloud

The traditional datacenter-centralized architecture collides with the latest trend of ubiquity of Cloud computing, as it confines the cloud computation to the location of the datacenters, generally far from the user. In this thesis, we focus on two drawbacks of datacenter-centralized Clouds: energy consumption and quality of service. On the one hand, due to its centralized nature, energy consumption in networks is affected by the traditional vision of the Cloud. That is, backbone networks increase their energy consumption in order to connect the rising number of clients with their growing traffic to the Cloud datacenters. On the other hand, distance leads to increased utilization of the broadband Wide Area Network and poor quality of service, especially for interactive applications. A distributed approach can provide a better quality of service and energy efficiency in large urban areas. To do so, it should confine local traffic close to the user and take into account the energy profile of network resources. In this thesis, we propose a novel distributed cloud architecture called microclouds. Microclouds are dynamically created and allow users to contribute resources from their computers, mobile and network devices to the cloud. This way, they provide a dynamic and scalable system without the need of an extra investment in infrastructure. This approach is evaluated on two realistic use-cases: a core network and a mobile cloud in a smart city. Through simulations, we show an important overall saving of energy consumed in core networks with our microcloud approach. Also, our results indicate that this microcloud architecture is scalable with the number of mobile devices in a smart city scenario, and provides a significantly lower latency than traditional datacenter-centralized approaches. Finally, we analyze the use of incentives for inciting users to contribute resources in our approach, and propose a new auction system adapted to the high dynamism and heterogeneity of these systems. We compare our solution to other existing auctions systems in a Mobile Cloud use case, and show the suitability of our solution.

A drink after the defense in room Sein

Thursday, 9. February 2017 - 10:00 to 12:00
IRISA-Inria salle Petri-Turing - campus de Beaulieu RENNES
Defense Type: 
Composition of jury: 
  • STOLF Patricia, Maître de conférence, Université Paul Sabatier, Toulouse, rapporteur
  • SENS Pierre, Professeur, Université Paris 6, Paris, rapporteur
  • URVOY-KELLER Guillaume, Professeur, Université Nice Sophia Antipolis, Nice, examinateur
  • PIERRE Guillaume, Professeur, Université Rennes 1, Rennes, examinateur
  • PHILLIPS Chris, Reader, Queen Mary University of London, Londres, GB, examinateur
  • LEBRE Adrien, Maître de conférence, Ecole des Mines de Nantes, Nantes, examinateur
  • MORIN Christine, Directrice de Recherche, Inria, Rennes, directrice de thèse
  • ORGERIE Anne-Cécile, Chercheur CNRS, Rennes, directrice de thèse