5G (NR) Network Analysis and Trigger Conditions

I. Network Analytics is a 5G system utilizing artificial intelligence/machine learning-driven real-time data analysis; it monitors and optimizes network performance, user experience, and resource allocation based on the 3GPP standardized NWDAF (Network Data Analytics Function). Network analytics achieves proactive closed-loop automation by collecting fine-grained data from the Radio Access Network (RAN), core network, and User Equipment (UE), thereby improving service quality, managing network slices, and predicting network behavior.

II. Network Analytics Features: Enabling network analytics provides mobile network operators with the following advantages:

• Increased Efficiency: Optimizing network resources and reducing total cost of ownership (TCO);
• User Experience Optimization: Monitoring and improving end-user quality of experience (QoE);
• Operations Optimization: Replacing passive manual troubleshooting with automated, proactive, and predictive operations;
• Vendor Interoperability: Using standardized interfaces to avoid vendor lock-in.

III. Key Network Analytics Nodes:

• NWDAF (Network Data Analytics Function): This is a core 5G function that collects data from multiple network nodes, generates and analyzes data, and provides insights to support automated operations.
• Fine-grained Real-time Data: Supports monitoring traffic at the user, session, and application levels to ensure high-quality service, especially for critical 5G services.
• Predictive and AI-driven: Utilizes machine learning to analyze historical and current data for proactive network management, such as predicting congestion or mobility issues.
• Automated Closed-loop: Enables the network to automatically adjust itself based on analytical insights without manual intervention.
• Network Slice Optimization: Provides specialized insights for managing the performance of different network slices, ensuring dedicated resources for specific services (e.g., high-bandwidth or ultra-low latency applications).

IV. Network Analytics Triggers: In the 5G system, the SMF requests or subscribes to analytical information from the NWDAF. The trigger conditions include the following conditions in the internal logic:

- UE PDU session-related events subscribed to by other NFs (e.g., AMF, NEF);

- UE access and mobility event reports from the AMF;

- Locally detected events;

- Received analytical information.

The trigger conditions may depend on the operator and SMF implementation strategy; when a trigger condition occurs, the SMF can decide whether any analytical information is needed; if needed, it requests or subscribes to analytical information from the NWDAF. When certain local events are detected, such as the number of PDU session establishments or releases within a specific area reaching a threshold, the SMF can request or subscribe to network analytics information related to "abnormal behavior" (as described in TS 23.288[86]) to detect any abnormal UE behavior within that area.