A novel customer-centric Methodology for Optimal Service Selection (MOSS) in a cloud environment

Cloud service selection decision has become tremendously challenging because of the exponential proliferation of cloud services. A judicious decision necessitates a thorough evaluation of services from sundry perspectives. While most existing studies evaluate services from the Quality of Service (QoS) perspective, they overlook the degree of delight or annoyance of a service user i.e. Quality of Experience (QoE). Likewise, the literature lacks an integrated methodology to (1) incorporate both QoS and QoE in decision making (2) develop a consensus between the contradictory outputs of Multicriteria Decision Making (MCDM) methods. To address these issues, we propose a novel integrated approach called Methodology for Optimal Service Selection (MOSS). MOSS consists of five stages including the prequel, assessment, ranking, integration, and consolidation/selection. MOSS enables decision-makers to select optimal cloud service with consensus considering both QoS and QoE. In the prequel stage, we introduce Pareto optimality to shrink search space and identify dominant services. In the assessment stage, we use the best worst method to calculate weights of QoS/QoE criteria. We employ a multi-MCDM approach consisting of eminent existing MCDM techniques to obtain QoS, and QoE based ranks in the ranking stage. We obtain and compare the integrated ranks of each method in the integration stage. We obtain the consolidated ranks of cloud services using the Copelands’ method. To verify the efficacy/practicability, we implement MOSS in the context of an e-commerce company facing a cloud service selection decision. Further, we perform a comprehensive analysis considering a comparative analysis and complexity analysis. The results show MOSS is practical and useful.