Virtualization CPE template automatic recommendation algorithm based on fuzzy comprehensive evaluation method

Research of virtualized Customer Premise Equipment (CPE) based on Network Function Virtualization (NFV) environment can replace existing physical CPE, thereby reducing the deployment cost. There are two main problems in research of existing virtualized CPE, one is that existing template configurations are manually configured, and configuration efficiency is low; the other is that the previous research on Virtualized Network Function (VNF) function templates for virtualized CPE mostly focuses on the research of a certain function or a specific scene, and lacks versatility. Aiming at these two problems, this paper presents a virtual CPE template automatic recommendation algorithm based on fuzzy comprehensive evaluation method. The automatic recommendation template is applicable to many different types of templates. Firstly, we classify the template parameters. Secondly, we build the index system. Thirdly, we calculate the weight assignment level of the membership index of the second-level index system based on the fuzzy comprehensive evaluation method. Finally, generate parameters. The model can automatically generate templates that are more suitable for current situation in real time according to changes in parameters such as traffic, user type, and device performance. Compared with manual configuration, there is a significant improvement in parameters generation accuracy and template configuration speed.