Reverse Auction to Trade Unused Cloud Computing Resources

Abstract

Objectives: This paper solves the resource underutilization issues by trading the unused cloud computing resources and also obtains an optimal profit with the help of reverse auction mechanism. Methods/Statistical Analysis: Cloud market uses direct selling model to allocate resources for fixed price. Cloud service providers are not considering marketing principles such as supply-demand rate. In existing works, several market-based resource allocation models and algorithms have been projected to show that dynamic pricing models might be more suitable for both the providers and the customers of cloud resources. The proposed work uses reverse auction mechanism to trade the unused resources along with the resource overbooking and pooling facilities. Findings: Cloud services are offered by more service providers. It is a complicated task to calculate the right cost for the usage by the cloud service providers. Reverse auction mechanism is used to trade the unused cloud computing resources, which is unable to be allocated through direct-selling. An adaptive bidding strategy is used to suggest the provider to take the right actions in auctions. This mechanism solves the issue of resource underutilization with the help of resource overbooking and resource pooling. Though overbooking is generally used to reduce penalty for the service providers, resource pooling facility is highly effective in accessing the resources from resource pool which can be utilized by service providers instead of paying penalty and also increases the datacenter utilization. Applications/Improvements: Resource pooling assisted reverse auction mechanism helps to improve the datacenter utilization and also reduces penalty for cloud service providers in both English Reverse (ER) and Second Price Sealed Bid (SPSB) auctions.