A Multiple Attribute Decision and Bidding Based Cloud Resource Dynamic Allocation Method

Cloud computing is on demand as it offers dynamic flexible resource allocation for reliable and guaranteed services in a pay-as-you-use manner to public. In cloud computing, a cloud resource consumer can request a number of cloud resources simultaneously. So there must be a provision that all resources are made available to requesting cloud resource consumers in an efficient manner to satisfy their need. In this paper, an intelligent combinatorial auction based resource allocation mechanism is proposed in a cloud environment. First, the system frameworks of resource allocation mechanism are constructed, and then tender descriptions are given for four kinds of common resources. Second, a support vector regression (SVM) based method is adopted to convert a combination resource demand into multiple single resource demands. Third, emotional parameter is introduced to bidding strategy and tender evaluation mechanism is designed to determine an optimal tender of cloud resource provider. Finally, a reputation system is introduced to deal with potential malicious behaviors in the auction market. Simulation results show that the mechanism is feasible and effective.