IP Tag Identification And Processing Technology For Next Generation Mobile Internet

In the wake of the information era,and with the popularization of 4G technology,the performance requirements of mobile Internet in terms of high speed,large capacity and business diversity are getting advanced to a higher and higher level.The next generation of mobile Internet and 5G technology are being closely deployed,how to satisfy multiple users coordinate a variety of network resources,to provide users better quality and fast service is the main goal of the operators and network platform in the high-speed connection environment.The identification and management of IP business tagging in an effective manner on one hand will classify data highly effectively,improve network efficiency,enhance the control of network,while on the other hand,will lay the foundation for the multi-service development planning for major operators to improve the quality of service and quality of experience.This topic is based on the Guangzhou Science and Technology Plan project that is for the study and industrialization of 5G network test cloud platform technology,proposes an IP tag recognition processing method for the next generation mobile Internet.In this paper firstly introduces the research background and development trend of business identification method under the mobile Internet,analyzes the principle,advantages and disadvantages of depth detection technology,emphatically introduces the decision tree algorithm based on machine learning and the retraining process of model training,and briefly describe the storm distributed processing engine mechanism for real-time data processing.Then,according to the three major application scenarios of the next generation mobile Internet,based on big data architecture and processing tools such as Hadoop,Storm and Netstat,the experimental environment which can enforce peculiarity of large capacity,multi-connection and high rate is constructed.Based on the experimental environment,the analysis of typical service protocols and data packets in instant messaging,video and knowledge sharing classes can be completed,and the decision node criteria for dividing different services can be achieved.It can provide a simple,feasible and universal experimental environment scenario for large-scale commercial use of 5G,and solve the problem of lack of test environment.Finally,based on the method of quality data set processing and combined with the improved C5.0 algorithm,Bagging algorithm framework,an enhanced classifier recognition scheme based on the quality data set is proposed.It can filter invalid data packets and enrich relevant data protocols to obtain high-quality data sets,so as to improve the identification efficiency of data sets and increase the identification accuracy of effective data packets.After that,the obtained high-quality data were divided and the corresponding classifiers were trained.According to each classifier to improve the network identification accuracy,that is conducive to the optimization and maintenance of the network,improve the operation performance of the service desk.Through the test of the scene environment,the verification of the experimental scheme and the comparison with other recognition schemes.The results show that,the experimental environment set up in this paper can realize the acquisition rate of 100 Gbps,support multiple connections,and have the storage capacity of TB level.The test environment provided in this paper can meet the requirements of the next generation mobile communication.IP tag recognition technology which based on high quality data set processing method and enhanced classifier training model improves the recognition accuracy of single classifier,increases the generalization ability of recognition and the overall recognition accuracy,and optimizes the operation performance of the workbench.