Research On Optimization Of Network Logical Topology Structure Of Access Network

With the rapid development of the Internet economy,when people enjoy the dividends of the Internet,they rely more and more on the Internet,and therefore have higher requirements on the Internet.With the development and deployment of 5G networks and the adoption of the NSA networking mode,some of the 4G network bandwidth has been squeezed,resulting in stalls and delays when users experience the network,and as the network traffic continues to increase,as well as exceeding part of the network The upper limit of the carrying capacity of meta-nodes faces the risk of local network element downtime at any time.In order to ensure the quality of the network and the safety of network elements,it is of great significance to optimize the network topology.The optimization of the network topology structure must not only rationally divert the traffic distribution in the network structure,that is,timely divert traffic in areas with excessive traffic,but also ensure the safe and stable operation of each network element node server.In addition,the optimization of the network topology can make the line operation of the communication network more reasonable and efficient through reasonable optimization methods under the premise of efficient operation.In addition,in the optimization of the network topology,the correlation between the core path and the backup path should be maintained at a low level.This paper proposes a staged optimization algorithm,which uses two different algorithms to optimize path selection and network element bearing conditions respectively.First,a hybrid algorithm based on Viterbi algorithm and particle swarm algorithm is used to establish a path planning model to solve the path planning problem under multi-objectives.Then a system recommendation algorithm model based on the mean Bayes algorithm is established,and the relevant evaluation indicators are converted into feedback functions,and solutions are recommended according to the conditions of different network elements through the model to achieve the goal of optimizing network elements.This optimization algorithm can achieve the effect of overall optimization through gradual local optimization.In the experimental verification,accuracy and score are used as the objective function.The experiment shows that the staged optimization algorithm proposed in this paper is better than the traditional optimization algorithm.Comprehensive analysis,this article uses a step-by-step optimization algorithm,compared with the traditional optimization algorithm,increased the flexibility of network topology optimization,the path planning is no longer limited to the limited forwarding path of network elements,so that the selectivity of the path is improved.Using a hybrid algorithm can maximize the goal of network performance optimization.In the network element optimization,it is ensured that the network element forwards the data in a balanced and efficient manner.The scoring of the experimental results proves that the optimization scheme proposed in this paper has a better optimization effect on the network topology in terms of comprehensive performance;it has high application value in the face of the optimization of the network topology.