Research And Simulation Of Network Slicing Algorithm Based On SDN/NFV

Due to the rapid and vigorous development of the Internet industry,a variety of businesses with different demands have emerged,such as the Internet of vehicles industry,telemedicine,smart home,5G industry,etc.Different businesses also have different demands for bandwidth,delay,storage,etc.The rigid traditional underlying network has been unable to flexibly meet the needs of these emerging network services,so the network slicing technology emerged as The Times require.With software-defined network(SDN)and network function virtualization(NFV)technologies,network slices can be quickly deployed and centrally managed,simplifying management,improving resource utilization and reducing costs.This article mainly studies the network slicing framework based on SDN/NFV and the core idea of SDN/NFV,and then studies the mapping method of network slicing for different problems.In order to solve the problem of long queuing time of high priority network slices in the process of network slice mapping,a priority-based shortest path greedy algorithm(PSPG)was proposed.Firstly,the network slice queuing model is modeled as a priority task queue model,and the queue is divided into two queues of high priority and low priority.Then,the mathematical model of the network slice mapping problem is established,and the optimization objective is defined as the network slice delay.Then,the network slice mapping problem is divided into node mapping sub-problem and link mapping sub-problem for solving.In the process of node mapping,the greedy algorithm is adopted to allocate resources with higher performance to network slices with higher priority,and the virtual network functions are mapped to corresponding devices.After the node mapping is completed,the shortest path algorithm is used for link mapping.Finally,the PSPG algorithm is simulated and the results are verified.In the process of network slicing mapping,time delay and resource occupancy are a pair of conflicting goals.Therefore,in order to solve the conflict problem of time delay minimization and resource utilization maximization in the process of network slicing mapping,a non-dominated sorting genetic algorithm(KMCNSGA-II)based on clustering is proposed to solve the mapping problem of network slicing.Since the optimization problem is a tradeoff between time delay and resources,a multi-objective optimization model is established,and the constraints in the network slice mapping process are defined.After KMCNSGA-II algorithm is used to complete node mapping,the shortest path algorithm is used for link mapping.Finally,KMCNSGA-II is used for experimental simulation and result analysis.