Intelligent O&M Solution for New-Generation Railway Mobile Communication System

The future railway mobile communication system (FRMCS) has entered an accelerated phase of development. Its mission to enable and facilitate digitalization will pave the way for train modernization, improving security, punctuality, services and capacity. In September 2023, China’s Ministry of Industry and Information Technology (MIIT) approved the test frequency for new-generation railway mobile communication system (5G-R), based on 5G technology, to China State Railway Group Co., Ltd. This approval marks the beginning of a new era of China's railway mobile communication.

With a new architecture, the 5G network fully supports three communication scenarios: ultra-reliable low-latency communication (URLLC), enhanced mobile broadband (eMBB), and massive machine-type communication (mMTC). It brings new experience to railway main-line train-to-ground communication, station/hub communication, infrastructure monitoring along railways,  and internal vehicle communication. However, the diversity of application scenarios and services presents greater challenges for network O&M and service quality assurance. Based on intrinsic intelligence, ZTE’s uSmartNet intelligent network solution focuses on  network collaborative intelligent O&M and service collaborative intelligent O&M at the NE layer, single-domain layer, and cross-domain layer through intelligent convergence, digital twins , and intent-driven approaches. This solution is essential for visualizing, managing, and controlling 5G-R networks and lays a solid foundation for intelligent network maintenance and service deployment in the future. The ZTE uSmarNet solution architecture is illustrated in Fig. 1.

Network Collaborative Intelligent O&M for E2E Closed-Loop Management of Network Events

The 5G-R network consists of the core network, radio access network, bearer network, and management system. Traditionally, O&M has been handled independently for each domain, leading to poor coordinate between low efficiency. For the new-generation railway mobile communication system FRMCS, a shift is needed from “equipment fault-oriented monitoring” to a sustainable, intelligent O&M mode focused on "network event-oriented management". This approach involves each domain providing alarm management and performance monitoring capabilities. By exposing these capabilities for upper-layer applications to invoke and integrate, a scenario-based end-to-end (E2E) event management O&M mode can be established. In addition, AI technology will be introduced to implement comprehensive intelligent O&M management.

Based on intelligent O&M technologies and application components, the E2E network event management implements policy orchestration and rule deployment tailored to each O&M scenario and service objective. It ensures closed-loop event management from perception and analysis to decision-making and execution. Key capabilities of the E2E network event management include:

• Key capability 1: Building an E2E closed-loop event management

The system offers comprehensive event lifecycle management, including event perception, analysis and diagnosis, decision-making, execution, and verification, with automatic and intelligent process handling. By collecting and reporting standardized data from each domain, the system identifies events through multi-source feature association, locates root causes based on the rule library and intelligent algorithms, and provides diagnostic insights and handling suggestions. For certain faults, the system can automatically orchestrate, schedule, deliver instructions, and implement self-healing.

• Key capability 2: Building AI integration to achieve intelligent O&M visualization

The NEs have intrinsic intelligence, with the AI engine embedded in network management. It supports multi-type data adaptation, and uses self-developed Adlik inference acceleration and automatic model optimization to integrate AI training, inference, and application. Through visual modeling and pipeline orchestration, event rules can be configured more flexibly, and multidimensional event billboards can be provided to visualize the management process.

Service Collaborative Intelligent O&M for Full Lifecycle Management of Service Quality

5G-R network service quality is crucial for ensuring safe and stable train operations. Oriented to 5G-R services, it focuses on four areas: indicator modeling, quality monitoring, problem tracing, and perception repair. Using solutions such as end-to-end problem delimiting, signaling backtracking and detail tracing, and one-click user exception delimiting, it develops capabilities for service quality evaluation, problem tracing, and closed-loop control, supporting the full lifecycle management of 5G-R network service quality and becoming the core of future O&M work. To achieve this goal, it is essential to focus on the following aspects:

• Service quality data collection is fundamental to service quality analysis and collaborative O&M. A a multi-dimensional data collection channel is established to collect, parse, integrate, and trace 5G-R control-plane, user-plane, and wireless interface data, and support signaling analysis across all interfaces and end-to-end service quality analysis.
• Service quality is monitored in real time, and faults are addressed promptly. By modeling network-layer, transport-layer, and service-layer KPIs and focusing on service development, network quality, and service perception, multi-granularity visual monitoring capabilities are established. This includes real-time control-plane guarantee and indicator alarm configuration across railway bureaus, marshalling yards, and DNNs, supporting real-time monitoring of network and service quality. Additionally, timely problem detection and immediate response are enabled to minimize impact.
• Cross-domain coordination and one-click delimitation make problem tracing easier. By utilizing unified modeling and data association, the single-domain self-intelligence capabilities of the 5G-R core network, wireless network, and bearer network are enhanced to enable root cause self-diagnosis and self-optimization of perception degradation. This approach further improves end-to-end cross-domain analysis and problem closed-loop management capabilities, effectively supporting service perception improvement. When a user service is abnormal, the one-click delimiting function can analyze end-to-end service perception problems based on the service ID, user ID, and occurrence time and location. It automatically outputs diagnostic results and provides optimization suggestions, so that the perception problem can be traced and handled easily.

Both network collaborative and service collaborative intelligent O&M have been applied in various 5G network scenarios, such as intelligently reducing orders from the alarm ticket system. By refining the alarm correlation knowledge map, alarm correlation rules, and network resource topology, the accuracy of fault delimiting and positioning is improved. This leads to effective online analysis and precise dispatch of alarm tickets. With an average number of 682 alarm tickets per day, intelligent analysis reduces 48 repeated tickets daily, achieving a 7% reduction rate. This fully proves the great potential of network collaborative intelligent O&M to improve work efficiency. Based on real-time service experience guarantees, dynamic identification of poor quality, real-time cross-domain analysis, and event closed-loop management, the unified O&M of 5G private networks has established a quality closed-loop management and control mechanism. This approach has been applied in the operator market to improve both service quality and efficiency.

The 5G-R system is a key infrastructure that will carry core services in the future, such as railway control, command, dispatch, and communication, ensuring safe and smooth railway operations. ZTE and industry partners are actively researching technologies and applications based on the intelligent development goals of “intelligent network operation, intelligent resource management, intelligent system maintenance, and intelligent service applications”. They aim to meet the development objectives of “visualized, manageable, controllable, measurable, reliable, and trustworthy" railway communications. Their commitment is to build a new generation of high-quality, highly reliable, and intelligent railway mobile communications system, advancing China’s railway mobile communications infrastructure.