基于多源数据的铁路5G-R运用质量评估系统设计

Design of Multi-source Based 5G-R Operation Quality Evaluation System

铁路5G-R网络性能监控和故障排查技术研究是提升5G-R运用水平的重要手段,针对各类网络性能数据获取困难、铁路管理和技术人员对于5G-R网络运用质量关注的颗粒度不同、智能化手段支撑不足、缺乏系统级网络性能评估体系设计等问题,探究如何充分利用多源数据建立5G-R网络性能智能化评估系统以支撑智能铁路发展成为迫切需要。深入分析DPI、Uu接口、网管北向接口、路测等用于评估的数据源的特点和构成,并介绍各类数据的获取方法。提出基于劣化度理论的5G-R总体健康状态判断思路,并利用机器学习技术对网络异常和故障进行分析。在以上关键技术研究的基础上,提出基于微服务的5G-R运用质量评估系统架构。针对5G-R运用质量评估开展的关键技术研究和系统架构设计可为未来5G-R运维支撑系统研发提供技术支持。

The research on 5G-R(5G-Railway) performance monitoring and troubleshooting technology is an important means to improve the level of 5G-R operation.In view of the difficulties in obtaining various network performance data,the different granularity of railway management and technical personnel's concern about the quality of 5G-R network,the insufficient support of intelligent means,and the lack of system level network performance evaluation system design,it is urgent to research how to make full use of multi-source data to establish an intelligent evaluation system for 5G-R network performance to support the development of intelligent railway.The characteristics and composition of the data sources used for evaluation are deeply analyzed,such as DPI,Uu interface,NMS north-bound interface,as well as drive test.The idea of judging the overall health status of 5G-R based on the theory of deterioration degree is proposed,and the network abnormalities and failures are analyzed by using machine learning technology.On the basis of the above key technology research,a micro-service-based 5G-R operation quality evaluation system architecture is proposed.The key technology research and system architecture design for 5G-R operation quality evaluation in this paper will provide technical support for the R&D of future 5G-R operation and maintenance support systems.