Research On Network Resource Management Method Based On SDN

As the core technology of 5G networks,network slicing technology can separate a unified physical network into multiple end-to-end virtual networks by networking on demand.network slices are logically separated.The core technology of network slicing is network function virtualization.Network function virtualization separates the hardware part and software part of the traditional network,the hardware is deployed by a unified server,and the software is undertaken by different network functions,so as to realize the requirements of flexible assembly services.However,due to the dynamic nature of the network traffic,the static allocation of resources in the network will lead to low resource utilization.Therefore,it is necessary to propose a more flexible resource allocation algorithm in order to meet a higher utilization rate of network resources,which is also the primary problem to be solved in this thesis.For the dynamic adjustment of network resources,firstly the traffic and allocated bandwidth in all tenant networks are monitored,and the bandwidth resource utilization is calculated separately.Secondly,each tenant network is marked according to the preset boundary value.Finally,the particle swarm optimization algorithm is used to obtain the best resource utilization,and the bandwidth resources are reallocated.In the process of allocating resources,the optimization function is designed to achieve optimal resource utilization.The simulation results show that the utilization rate of bandwidth resources can be improved obviously by particle swarm optimization algorithm.For virtual network mapping,this thesis presents a virtual network mapping algorithm based on ant colony algorithm.The residual resource is taken as the objective function and the ant colony algorithm is used for optimization,the updating formula of pheromone is optimized in the mapping process.In the node mapping stage,the virtual node maps to the physical node according to the mapping probability,while in the link mapping stage,the path with the smallest number of hops between corresponding physical nodes is selected,which improves the utilization rate of link resources.The simulation results show that the virtual network mapping algorithm proposed in this thesis brings good benefits.