Research On Data Center Network Traffic Classification Technology Based On Deep Learning

With the continuous development and application of cloud computing and big data technology,the scale of the data center network continues to expand and network traffic also continues to increase.The flow characteristics have changed from traditional north-south flow to east-west flow.The increasing traffic within the data center network may cause problems such as network link congestion and increased packet loss rate,which will affect the network performance of the data center.Therefore,the research and analysis of the data center network traffic is of great significance to improve the network transmission efficiency and quality of service.The paper studies data center network traffic from two aspects: traffic classification and Qo S routing algorithm.This paper studies the data center network traffic classification method based on deep learning.In view of the low accuracy of traditional traffic classification methods,a traffic classification method based on improved convolutional neural network is proposed.First,the network structure,network parameters,cost function,and data preprocessing methods of the convolutional neural network are optimized and improved,and then the improved traffic classification method is compared and analyzed with the existing classification method.The results show that the method of the improved traffic classification improves the classification accuracy of data center network traffic.This paper studies the Qo S routing algorithm of data center network based on SDN.In view of the difficulties of link state collection and low link utilization when data flows choosing the routing scheme in the data center network,a bandwidth constrained Qo S routing algorithm is proposed.Then the simulation experiment of the proposed algorithm is carried out in the simulation environment built by Floodlight and Mininet,and the effectiveness of the proposed algorithm is verified.Finally,based on the above research,this paper designs a traffic classification and routing system based on SDN,which consists of two parts: traffic classification and routing.