Research On Service Model And User Quality Of Experience For 5G Intelligent Railway

Intelligent railway is the inevitable trend and important task of railway development in China.The development of intelligent railway will hatch a series of new service types and service requirements.However,the existing private mobile communication system of railway cannot meet the intelligent demand of new services.With its powerful performance advantages,5G technology will become an important technical thrust for railway private mobile communication system,enabling business development and improving service quality.In order to cope with the challenges of diversified network users and refined service differentiation,mobile network management is gradually shifting from technology-driven mode to service-driven mode.Therefore,taking intelligent railway service as the core,this paper explored the application mode and applicability of 5G in intelligent railway scenario,excavated new intelligent railway business driven by 5G,studied the carrying capacity of 5G-R for intelligent business,and analyzed user satisfaction of service quality,so as to realize the collaborative optimization of network and business,at the same time meet the increasingly customized and differentiated business service requirements of the intelligent railway industry.Specifically,the main work of this paper includes:(1)From the perspective of network,studied the carrying capacity of 5G-R for services.The peak rate of 5G-R at 2100 MHz test frequency is analyzed and verified from theoretical calculation and practical test.Through simulated calculation and system test,it is proved that 5G-R can meet the load carrying requirements of intelligent railway service.(2)From the perspective of service,the intelligent railway business model is studied.By sorting out the demand indicators of traditional railway business and new intelligent business,the demand model of intelligent railway business is established.Based on the business demand,we used decision tree to build the business classification model,and the data enhancement methods of SMOTE,Borderline-SMOTE and ADASYN are used to solve the problem of small and unbalanced business data set.According to the classification results,the classification accuracy increased by 7%,9% and 13%respectively.(3)From the perspective of users,studied the experienced quality of the service.Taking railway video service as an example,the paper explored the factors affecting video Qo E,and carried out verification experiments in a real network environment based on5G-R video surveillance system.The user’s subjective experience was decomposed into three dimensions of video smoothness,clarity and real-time,and a hierarchical weighted Qo E mapping model was established considering the video Qo E impact factor.Fuzzy analytic hierarchy process is used to calculate the weight of each index to achieve the comprehensive evaluation of user quality of experience.