Adaptive Diversity of Genetic Algorithm for QoS-based Cloud Service Selection

摘要

An elitist selection adaptive Genetic Algorithm is presented to select an optimal cloud service composite plan from a lot of composite plans on the basis of global Quality-of-Service (QoS) constraints. In this Algorithm, some strategies are provided to improve convergence and population diversity and raise fitness result. These strategies include an elitist selection strategy, an individual selection method based on ranking selection and an adaptive mutation strategy. The elitist selection strategy gives a guarantee that the degradation of population fitness can be avoided. The individual selection method provides a simple and effective way to control selective pressure. The adaptive mutation strategy uses a special population measurement to decide individuals' similarity. The value of the dynamic mutation probability changes on the basis of population diversity. Some simulation results on cloud service selection with global QoS constraints have shown that the elitist selection adaptive Genetic Algorithm can gain better composition service plan.