Efficient dynamic resource provisioning based on credibility in cloud computing

Abstract

Cloud Computing is a growing technology in which resources are provided as a service. The efficiency in providing the resources as a service depends on various factors. One of the major concerns is the suitable allocation of resources to the job. Hence, this paper introduces an Auction based model (CMCDA) for selecting the best customer-providers pairs based on credibility for resource provisioning in cloud computing environment. The Credibility based Multi-attribute Combinative Double Auction (CMCDA) model reduces the complexity in providing the resources for the execution of jobs and fulfill the expectations of both the customers and providers in cloud computing environment. The model also finds the best customer-providers pairs in an efficient way by calculating the credibility values before the resource provision. The highest credibility values pairs are selected as the best customer-providers pairs in the list. Here, the credibility value represents the level of customers and providers satisfaction. The time complexity of the proposed CMCDA algorithm is O(nlog(n)), Since the algorithm only goes through the sorted bid and tries to match them, being executed at most l(n + m) times. The performance of the proposed CMCDA is compared with the Combinatorial Double Auction Resource Allocation (CDARA) model which is the existing cloud double auction model, using CloudAuction simulator. The experimental results demonstrate that the proposed CMCDA performs efficiently than the existing CDARA model for resource provisioning in cloud environment.