Research On Service Traffic Analysis And Resource Management Technology Of 5G

5G communication has developed more complex and diverse application scenarios based on 4G communication: Enhanced mobile broadband scenarios require extremely high data rate;massive machine type communication requires dense device connectivity;low-latency and high-reliability communication scenarios have strict requirements on the two indicators of delay and reliability.In order to meet the communication requirements of these three application scenarios,software defined network and network slicing have become key enabling technologies for 5G.Network slicing divides the infrastructure network into multiple end-to-end logical networks to provide services according to different business needs.Software defined network technology brings network programmability and reduces hardware upgrade cost.All along,study on service requirements and traffic models has received extensive attention from research organizations and academia in order to estimate communication needs and network planning predictively.The high mobility of vehicles in the eV2X system will lead to frequent handovers among cells,therefore reducing or avoiding service interruption while meeting service requirements is an important issue when researching mobility management.However,the traditional service models and traffic models are no longer applicable to 5G network with different service requirements and diverse scenario characteristics.The handover method of selecting target cells according to RSSI cannot meet the delay and mobility required by the eV2X system.Therefore,this paper did some research mainly on traffic analysis of uRLLC and eV2X handover method.The thesis firstly explains the research background and research status at home and abroad of traffic models and handover methods in eV2X system as well as the content and research basis of this article in the introduction chapter.URLLC require extremely high reliability and low latency.The service modeling and scenario traffic analysis will provide important basis for traffic control and network planning in 5G network.This paper uses a two-layer model to model the session and grouping layers of uRLLC services.Based on different business characteristics,different service models are established,and the probability density function and cumulative distribution function of data transmission rate are derived based on the established service models.The analytical results of the data transmission rate and the statistical results obtained by the Monte Carlo method are compared so as to prove the correctness of the service model.Based on the research results of service model,traffic model of uRLLC is deduced by using the traffic convolution principle of multi-user and multi-service scenarios.Then the data transmission rate and the scenario traffic are further calculated and analyzed to obtain the traffic characteristics of a single scenario,which helps network simulation in academic research and network laying in practical applications.In the eV2X system,vehicles with high mobility have services with low latency,high reliability,or high data rate.Therefore,to ensure service continuity,it is necessary to study mobility management methods suitable for 5G network architecture.This paper considers the eV2X network slicing architecture and proposes a prediction and network slicing based handover method: AMF predicts parameters of UE behavior model in advance to provide reference data for upcoming handover decisions,therefore reducing service transmission delays.In addition,when selecting the target g NB and network slice,AMF and NSSF consider the load,Qo S requirements,and mobility requirements in order to select the most suitable network for the UE.By simulating the handover algorithm and analyzing its performance,it can be seen that handover method based on the network slicing technology and prediction,which has achieved a certain degree of performance improvement compared to existing handover methods,can meet the Qo S requirements of eV2X services.