Research And Implementation Of Fast Virtual Network Embedding Algorithm

Nowadays,as the increasing network resource requirements coming from the new technology,the substrate network scale provided by Infrastructure Providers and the virtual network scale required by Service Providers are expanding resulting in the mapping time growth of the virtual network embedding(VNE).Meanwhile,with the development of the edge computing,kinds of applications need to face the real-time scene in which the VNRs change frequently causing continuous mappings.Thus,the embedding speed of the VNE algorithms should be as fast as possible to meet the demands from above.Yet,it is difficult to both keep the embedding time and the acceptance ratio reasonable in a large-scale network at the same time for most of the present VNE algorithms.This thesis proposes a fast virtual network embedding algorithm,called KCS-VNE,based on a resource-aware K-Core method called VNEK-Core which helps to extract the well-resourced sub-network in the original substrate network.The KCS-VNE is designed to accelerate the embedding speed of the virtual network embedding by extracting the well-resourced sub-graph in substrate network and guarantee the other metrics depending on the hierarchical feature of the K-Core at the same time.What's more,by introducing two metrics,the parameters used in KCS-VNE are well tuned.Varies of simulation experiments show that the execution time of the proposed algorithms is hugely cut down compared with other VNE algorithms reproduced in this thesis.In addition,our algorithm can also guarantee the acceptance ratio and R/C at the same time.Besides,this thesis also introduces a virtual network embedding prototype system who can simulate both the static and the dynamic virtual network embedding experiments.It can also generate user-defined network topologies for the certain experiment.According to the testing,this prototype system is easy to use and can show users a clear embedding result.