Follow the renewable-based virtual machine migration among geo-dispersed modular data centers with expert systems

  1. SEDDIKI BEN FRIHA, DORAID
Supervised by:
  1. Sebastián García Galán Director
  2. Rocío Josefina Pérez De Prado Co-director

Defence university: Universidad de Jaén

Defense date: 30 September 2024

Committee:
  1. Juan Ramón Velasco Pérez Chair
  2. Carmen Martínez García Secretary
  3. Luis Magdalena Layos Committee member
Universidad: University of Jaén

Type: Thesis

Teseo: O868107 DIALNET HANDLE: https://hdl.handle.net/10953/3396
RUJA. Repositorio Institucional de la Producción Científica de la Universidad de Jaén: lock_openOpen access Handle

Abstract

This PhD examines the issue of energy consumption in the cloud computing industry, driven by the growth of data centers (CDC) from companies like Google, Amazon, and Microsoft. These CDCs, consuming over 2.5% of the world's electricity and contributing 2% of carbon emissions, require strategies for greater efficiency and sustainability. The thesis proposes integrating renewable energy sources and inter-cloud planning techniques to optimize energy consumption. It suggests using fuzzy expert systems to manage uncertainty and enable efficient virtual machine (VM) migration between CDCs based on renewable energy availability. Additionally, knowledge acquisition mechanisms and bio- inspired AI techniques are implemented to enhance the efficiency of these planners. The proposed system's effectiveness is evaluated through simulations with WorkflowSim-DVFS, aiming to create a more environmentally friendly cloud computing industry.