Loading...

Analysis of Resource Allocation in Federated Clouds Using Game Theoretic Approach

Mohammad Mehdi Pour, Fauteme | 2019

389 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 52014 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Izadi, Mohammad
  7. Abstract:
  8. While Cloud providers face problems during providing services to their clients, such as lack of resources or increasing idle resources in low workload; they can share their resources with other providers by participating in Cloud Federations, and take advantage of Cloud Computing benefits like high scalability and resource extension. Moreover collaborating Cloud Federations expands Geographical area of providers. The Resource Allocation Problem has been considered in many researches, but in fields like Cloud environment and Federated Clouds it comes with complications. This thesis analyzes an optimized mechanism based on game theoretic approach for Resource Allocation Problem in Federated Clouds.Members of the federation in this game are considered infrastructure service providers, who are trying to attend the federation, to extend or sell their resources. These providers can appear as both service providers and clients. In this thesis, a client provider acts as the leader of a Stackelberg Game, to buy its required resources from the follower providers. This service provider leads the game in a way that maximizes its own profit, by determining the price of one unit of resource in every iteration of the game.In the implementation of this issue we take advantage of Cloudsim simulator, which is a powerful tool in cloud simulations. Results obtained from the Stackelberg Game algorithm are compared with the resource allocation default algorithm of the simulator, and improved it as desired
  9. Keywords:
  10. Federated Cloud ; Game Theory ; Cooperative Game Theory ; Stackelberg Game ; Resources Allocation

 Digital Object List

 Bookmark

No TOC