An Evolutive Hybrid Approach to Cloud Computing Provider Selection

• Main Author: Borges De Moraes L.*
• Publication Date: 2018
• Other Authors: Fiorese, Adriano, Parpinelli, Rafael Stubs
• Format: Conference object
• Language: eng
• Source: Repositório Institucional da Udesc
• Download full: https://repositorio.udesc.br/handle/UDESC/6116

© 2018 IEEE.The success of cloud computing technology has leveraged the emergence of a large number of new companies providing cloud computing services. Choosing which cloud providers are the most suitable to attend consumers desired quality of service has become a hard problem. In order to qualify such providers, performance indicators (PIs) are useful tools for systematic and synthesized information collection. Thus, the problem approached in this work is to find the best set of cloud computing providers that satisfies a customer's request, with the least amount of providers and the lowest price. Hence, this work proposes a hybrid Genetic Algorithm (GA) to address this problem. In experiments, three approaches, using PIs as input, are employed: A simple matching algorithm, a GA and the proposed hybrid matching-GA approach. The hybrid method combines the qualities of both the matching algorithm and the GA showing promising results.