Design And Implementation Of Virtual Network Mapping Method Based On SDN

With the development of cloud computing and big data technology,gradually formed a cloud platform, cloud platform to provide services for multi-tenant, in order to customize the network service can be flexible,the gradual emergence of software defined network, it has the forwarding and control characteristics in isolation, control and network virtualization.The network is used to share and isolate the underlying network, the virtual network requests are mapped to the underlying network. In the virtual network mapping process, the underlying network of key nodes or key path is priority mapping, easy to cause the bottleneck problems in the process of mapping, the research has an important significance to improve the utilization rate of the underlying network operators.This paper summarizes the basic theory and research status,characteristics of SDN network virtual network mapping method,analyzes the key resources based on the existing topology aware mapping method is the priority and become a bottleneck problem, proposes the virtual network mapping method is applied in SDN network based on topology aware to solve the problem. In this method firstly, through resource method using topology aware, dividing the underlying resource,when mapping, in order to adjust the resources according to the connectivity constraints, so as to avoid the key resources and priority allocation has become a bottleneck problem. Secondly, the optimization of dynamic reconfiguration, monitoring the bottom load, when the threshold is exceeded, the state through the detection of virtual network topology aware method, using the method of path migration and reconfiguration are adjusted to avoid the dynamic mapping of the resource in the process of load imbalance situation. Finally, the design and development of the application platform, the method is verified by simulation. The simulation results show that the proposed SDN based virtual network mapping method can effectively improve the virtual network request acceptance rate, the cost to income ratio and the load balancing ability of the underlying resources.