Loading...

Providing a Multi-Attribute Combinatorial Double Auction Model for Resource Allocation in the Cloud Environmet

Hassanzadeh, Reihaneh | 2016

615 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 48708 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Movaghar Rahimabadi, Ali
  7. Abstract:
  8. In auction-based cloud investment environments, consumers desire the least possible cost and high satisfaction, while providers seek strategies to maximize profit and return on investment. In these environments designing an adequate resource allocation that can satisfy both consumers and providers is of utmost importance. Yet, most available models in resource allocation focus on providers’ needs and do not concern themselves with the consumers. Therefore the present study aims to present an efficient and optimum model for resource allocation problem that takes into account various factors, e.g. resource price, service quality, reputation and fairness, and is capable of satisfying all honest participants. Normally the algorithm only cares for the overall profit of all participants, but adding fairness means that the algorithm assigns some value to each individual’s profit as well and tries to reduce the gap between poor and rich bidders. Furthermore, policies such as punishment and including reputation were adopted in order to deal with dishonest individuals that try to win an auction by providing false information. For resource allocation strategy purposes, the proposed method utilizes a linear programming algorithm and uses it for selection of auction bidders and resource allocation. Although the main goal of the algorithm is to maximize profits, it also tries to improve participant satisfaction by including more factors, e.g. service quality, reputation and fairness. Therefore, the proposed model tries to strike a balance between profit maximization and user satisfaction. Simulation results show that adding fairness, reputation and service quality attributes to the algorithm result in higher users in future auctions and prevent potential algorithm abuses
  9. Keywords:
  10. Resources Allocation ; Cloud Computing ; Reputation ; Service Quality ; Combinatorial Auction ; Justice

 Digital Object List

 Bookmark

No TOC