A Novel QoS Prediction Approach for Cloud Service Based on Bayesian Networks Model
Pengcheng Zhang, Qing Tian Han, Wenrui Li, HARETON K. N. LEUNG, Wei Song · 2016
Considered as the next generation computing model, cloud computing plays an important role in scientific and commercial computing and draws wide attention from both academia and industry. In the dynamic, complex and changeable cloud computing environment, Quality Of Service (QoS) is an important basis for the selection of different cloud services. Therefore, the prediction of cloud services QoS can help users to choose the most suitable service at hand. The software and hardware and resources of three-layer structure for cloud computing will impact on cloud services QoS, but existing QoS prediction approaches are not consider the three-layer structure on the influence of thecloud service QoS. The CPU usage, physical memory usage andthe number of processes of infrastructure layer have definitely influenced QoS. In order to address this limitation, in the paper, a Bayesian network model of QoS prediction for cloud servicesis proposed. Firstly, an initial and basic Bayesian network modelis established by collecting data from the infrastructure layer, the platform layer and the application layer. Then the Bayesian network is trained and updated to obtain the cloud service QoS prediction model. Finally, a set of experiments based on collected data from the real cloud service environment has been conducted to validate the proposed approach. Experimental results show that the prediction approach is effective and accurate.