Federated Architecture for Ranking the Services in Cloud Computing

摘要

Objectives: To improve Quality of Services (QoS) and ranking cloud service providers in federated cloud environment. It is also aimed to resolve various issues faced by the user and providers in cloud. Methods: In order to avoid the workload of existing cloud model and to compete with the well-known cloud service providers, customized federated cloud provider architecture was suggested. It is suggested to manage and to maintain QoS for the submitted tasks. The proposed architecture consists of external world, middle world and internal world. Each world plays vital role for customizing the request in federated environment to enhance QoS. To avoid starvation in the architecture to suggest the differentiated module and resolves the various key issues using the Stochastic Markov process model and also evaluating the cloud providers based on the quality of service requirements. Findings: Service Level Agreement (SLA) in single and federated cloud plays a vital role in enhancing Quality of Services. Single cloud service provider representation does not provide the skilled QoS, when workload becomes high. Customized federated cloud architecture reduces the drawbacks of single cloud service provider. The proposed architecture was implemented in CloudSim using Java. The simulation result of the proposed architecture proves enhanced QoS to the user and cloud service providers. The proposed federated cloud model enhances the Quality of Services by more than 18% of existing single and federated cloud model. The parameters considered for the simulation for numbers of users, providers, load factor, turnaround time and average load deviation of tasks. Applications: This architecture can be used to rank different cloud service providers and it can be trusted for any distributed cloud services with extended QoS.