Validation And Verification Platform Of SDN Controller Placement And Incremental Capacity Optimization

Software defined network(SDN)is a new kind of network architecture,which separates the control plane and the data plane.SDN data plane is composed of switches,which is responsible for the packet forwarding function of data streams.SDN control plane is composed of controllers,which is in charge of the centralized management and control of the data plane.In SDN network,it usually needs to rationally design the placement of the controllers in the network and the allocation strategy of the service resources to obtain profitable network performance.Therefore,the network design needs to be verified in advance.However,because the controller placement problem is NP-hard,the scenarios involved in network design increase exponentially,which makes it difficult to verify the network performance in different design scenarios.Therefore,there is an urgent need for related tools to provide guidance for the optimal deployment of the controllers and the optimal utilization of the resources.In addition,since network traffic with self-similar characteristics is considered to be widely present in modern communication networks,the use of traditional Poisson processes to describe network traffic models may cause deviations in the evaluation results.Therefore,in the research of the scientific problem of SDN resource optimization,it is also necessary to establish different network traffic models to analyze the problem based on both self-similar network traffic and network traffic arriving at Poisson.To this end,based on the research needs of some scientific issues of resource optimization and performance verification in NSFC,the paper studies related scientific issues and then designs and implements a validation and verification platform based on different network traffic models to verify the network performance boundary of SDN controller placement and incremental capacity problems.The paper combines the network traffic with the service capability of the controlle-rs,models the performance of the SDN controller based on different network traffic models and analyzes the controller queuing delay,and then combines the queuing delay and the propagation delay to establish the performance optimization indicator of SDN controller placement and incremental capacity problems.On this basis,the verification platform is designed and implemented.Finally,by setting different experimental parameters in the platform,the problems of controller placement and incremental capacity are verified through different experiments.In addition,some experimental conclusions are obtained.The verification platform in the paper can verify the performance boundary of the SDN controller placement and the performance boundary of the incremental capacity problem based on different network traffic simulation,and provide the optimal deploy-ment solution and the optimal capacity incremental solution for the controller.At the same time,the paper proves the importance of optimized deployment of controllers and optimized utilization of controller capacity incremental resources through related expe-riments,and analyzes the characteristics of the network traffic.Starting from the verification problem of network design,the paper researches various network traffic models and the controller performance modeling,and it uses network delay as the performance optimization index for controller placement and incr-emental capacity.On this basis,the validation and verification platform of SDN contro-ller placement and incremental capacity optimization is designed and implemented,which provides references for the research and practical application of some scientific issues of resource optimization and performance verification in software defined networks.