Abstract

Cloud providers can form cloud federations by pooling their resources together to balance their loads, reduce their costs, and manage demand spikes. However, forming cloud federations is a challenging problem, especially when considering the incentives of the cloud providers making their own decisions to participate in cloud federations. In this paper, we model the formation of cloud federations necessary to provide resources to execute Map-heavy/Reduce-heavy programs while considering the trust and reputation among the participating cloud providers. The objective is to form cloud federations with highly reputable cloud providers that achieve maximum profit for their participation. This is an NP-hard bicriteria optimization problem. We introduce a coalitional graph game, called trust-aware cloud federation formation game, to model the cooperation among cloud providers. We design a mechanism for cloud federation formation that enables the cloud providers with high reputation to organize into federations reducing their costs. Our proposed mechanism guarantees the highest profits for the participating cloud providers in the federations, and ensures high reliability of the formed federations in executing the applications. We perform extensive experiments to characterize the properties of the proposed mechanism. The results show that our proposed mechanism produces Pareto optimal and stable cloud federations that not only guarantee that the participating cloud providers have high reputation, but also high individual profits.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call