Priority based clustering the service providers in federated cloud

Abstract

Cloud is a trending technology now a days. In this technology world now everyone are moving with the cloud. A cloud federation technique is one of the best ways to select any services from the service providers by the user at any time. As large number of service providers in the federation, user need to be select best service provider and the selection of service provider get delayed in the federation. We proposed the cluster based approach to select the best service provider for user request and reducing the selection time of service provider. In previous papers there is no Automated Selection for the selection of best service provider. Our proposed work is based on Automated Selection model i.e., we applied k-means algorithm for grouping the similar services and categorizing the service providers into three groups. Such as Cluster-1, Cluster -2 and Cluster -3, after cluster these cluster sorted in order based on the final centroid of each cluster. Now user request to be compared with the middle cluster centroid and directed to respective cluster for service selection. If more than one service provider offers similar services in the cluster then priority decision tree to be invoked for tie break. This technique works better than existing algorithms and time also to be saved in selection of service.