Research On Transmission Service Model And Resource Management Technology Of 5G Network

5G system will support various scenarios of vertical industries and multiple types of services with different Quality of Service requirements.Among them,massive Machine Type Communication(m MTC)is an emerging type of industry application as one of the three major scenarios of 5G which needs to ensure the data transmission of a large number of connected devices to achieve the vision of Internet of Things.5G system needs to meet performance requirements which are much higher than 4G system to support explosively increasing data traffic and massively connected devices.At the same time,the multiple types of services,high traffic density,and high device connection density also bring challenges to the management of network loads.This article focuses on services traffic and load balancing,including research on the mathematical modeling of m MTC services,traffic statistics for m MTC scenarios and mixed scenarios,and inter-cell load balancing based on user's mobility prediction.This paper first studies the existing service model.Taking into account the difference between machine type communication services and traditional human-to-human communication services,this paper uses a general simplified two-layer model to describe the m MTC services from the session layer and the packet layer.This paper then selects household energy management systems,smart grids,wireless vending,home monitoring,shared bicycles,automotive applications,and other typical services,establishes mathematical models for simulation,and compares with Monte-Carlo simulation to analyze the data rate characteristics of service and verify the accuracy of the service model.Based on the research foundation of single-service modeling,this article continues to study traffic statistics in different scenarios.For multi-user and multi-service scenarios,this paper establishes a convolution-based traffic model and describes statistical methods.Then,for m MTC scenarios such as smart houses,smart buildings,and public services and for mixed scenarios such as smart parks,according to analytical and Monte-Carlo simulation comparison,this paper analyses the traffic characteristics of thesse scenarios.Finally,the approximate eighth-order Gaussian distribution of the traffic model is obtained by the method of curve fitting.The traffic models established in this paper have certain versatility,which can provide a reference for planning of communication network and for research of related technology.Finally,based on the 3GPP network architecture,this paper studies the prediction of user's mobility in cellular systems.This paper firstly describes the general prediction method based on historical locations of users,and then selects support vector machine model and neural network model to predict user's mobility based on self-similar least-action human walk model and comparative analysis is made.Based on the prediction results,this paper balances the load between cells through resource reservation and access control methods.On the premise of ensuring the user's Quality of Service requirements,the user selects the appropriate cell to access according to the cell load and rate requirement of usr'e service.Simulation results show that this method achieves a reasonable compromise between satisfying the quality of service and balancing the system load and has certain increase in resource utilization.