Modélisation et optimisation de réseau infonuagique et de calcul périphérique

Abstract

Edge computing has recently emerged to provide services for applications with strict constraints in terms of quality of service, such as low-latency 5G applications. However, due to their limited computing capacity, edge data centers are not able to serve a large number of user requests. A combination of edge and cloud is therefore required to provide scalable service closer to end users. The main challenge faced by such a complex system including edge, core data center and public clouds is optimizing resource allocation to guarantee Quality of Service (QoS) requirements and to meet capacity constraints. In this thesis, we present a mathematical model to optimize resources allocation in edge-cloud environment for multiple applications simultaneously. Traditional optimization methods are no longer appropriate for solving this problem. We model the problem as a non-cooperative game and show that it can be transformed to a classical transportation game. We propose a solution to the game based on the elimination of strictly dominated strategies and the best response method. The goal is to identify the Nash equilibrium among equilibria set generated in the game. Experimental simulations show our solution outperforms the prior work.

Date	11 May 2021
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Kim Khoa Nguyen (Supervisor) & Mohamed Cheriet (Co-supervisor)