Research On Some Key Technologies Of 5G Networks Digital Twin Model

With the development and wide application of mobile network technology,wireless networks not only need to meet the use of massive mobile data and high-speed network speeds,but also improve service quality.Therefore,efficient network management and control is a huge problem.The management and control of traditional networks requires a lot of manpower and material resources,and efficiency is also an issue.Digital twin technology enables physical space and virtual space to interact with each other,comprehensively utilizes technologies such as perception,operation,and model,and uses software definition to realize description,diagnosis,prediction,and decision-making of physical space.It can provide powerful technical support for the management and control of modern networks.In view of the new problems introduced by these technologies,this thesis deeply studies the network fault diagnosis and service type identification based on the 5G network digital twin model,which provides new ideas for the introduction of these new technologies.Based on the introduction of future network development trends and existing network fault diagnosis methods and network service type identification methods,this thesis mainly conducts the following three aspects of research:(1)Fault detection and diagnosis based on the 5G network digital twin model method.First,build a digital twin network model for the wireless access network and core network of the 5G network,including network elements,relationships between network elements,and functional unit models.Then establish a network fault diagnosis model based on digital twins,including data processing and knowledge models.For fault diagnosis,analyze the key parameters and network fault categories,train the knowledge model based on the data set composed of the actual collected data and the simulation data of the digital twin platform,and perform network fault detection,diagnosis and prediction based on the knowledge model.The accuracy rate of network fault diagnosis reaches 98.82%.(2)Network service type analysis method based on 5G network digital twin model.First,analyze in detail various common business types.Then introduce common service type analysis methods,including port number-based service type analysis methods,and deep packet inspection(DPI)-based service type analysis methods.Finally,in order to solve the shortcomings of the above methods,a detailed analysis of the business type analysis method based on machine learning,the knowledge model of the digital twin network-SVM model and the random forest model are applied to the actual business type analysis,including the actual network data The data collected by the packet collection platform and the simulation data generated by the digital twin platform are used to identify the business type,so that the recognition accuracy rate reaches 97.3%.(3)Visualization platform for fault diagnosis based on digital twin network.Using the existing data and information,a digital twin network fault diagnosis visualization platform is built with Javascript,through which the fault diagnosis situation of each region is displayed in real time,and the accuracy of the fault diagnosis prediction model is displayed.