Priority-Based and Optimized Data Center Selection in Cloud Computing

Abstract

Cloud computing (CC) is rapidly emerging in IT fields and becomes a vital research area. CC environment allows cloud service providers to offer different types of services to customers of different categories. Customers like their requests to be processed at the lowest cost and minimum response time, which is highly dependent on the receiving datacenter (DC). Selecting a DC is entitled to the service broker that operates in accordance with a built-in routing policy. Improper selection of a DC may cause a bottleneck in the service brokerage architecture, resulting in resource bleeding (time and money). Moreover, cloud services are charged based on a pay-per-use model. Therefore, this paper proposes an efficient service broker routing policy that improves users’ satisfaction and cloud performance. The proposed policy employs the Vector Space Model and a multi-objective scalarization function to optimize conflicting objectives. The simulation-based approach was used to test the performance of the proposed methodology. The results show the ability of the proposed methodology to improve DC selection process based on the user’s priorities. Moreover, the performance of the proposed methodology has been compared with related works; the results show a significant improvement in average DC processing time, response time and total cost.