A Multi-Dimensional Fairness Combinatorial Double-Sided Auction Model in Cloud Environment

TL;DR

This paper introduces a multi-dimensional fairness combinatorial double auction model for cloud resource allocation that balances profit and fairness, improving participant engagement and resource utilization.

Contribution

It proposes a novel auction model incorporating fairness based on participant history, addressing limitations of existing profit-focused models.

Findings

Increases participant willingness to join future auctions

Enhances average resource utilization

Balances profit and fairness effectively

Abstract

In cloud investment markets, consumers are looking for the lowest cost and a desirable fairness while providers are looking for strategies to achieve the highest possible profit and return. Most existing models for auction-based resource allocation in cloud environments only consider the overall profit increase and ignore the profit of each participant individually or the difference between the rich and the poor participants. This paper proposes a multi-dimensional fairness combinatorial double auction (MDFCDA) model which strikes a balance between the revenue and the fairness among participants. We solve a winner determination problem (WDP) through integer programming which incorporates the fairness attribute based on the history of participants which is stored in a repository. Our evaluation results show that the proposed model increases the willingness of participants to take part in the next auction rounds. Moreover, the average percentage of resource utilization is increased.