NWDAF

Technical Explanation

The NWDAF represents 3GPP's native standardization of AI/ML within 5G. It collects real-time event logs, metrics, and telemetry from NFs (AMF, SMF, PCF), OAM (Operations, Administration, Maintenance), and MDT (Minimization of Drive Tests). NWDAF applies statistical algorithms and trained ML models to generate analytics and predictions for slice load level, abnormal UE behavior, user data congestion, mobility predictions, and QoS sustainability, feeding actionable insights back to PCF, NSSF, and AMF.

Key Functions

• Telemetry data collection from 5GC NFs, OAM, and UEs
• Machine Learning inference for network slice load forecasting and optimization
• Predictive mobility and communication congestion analytics
• Providing automated inputs to PCF for dynamic policy adjustments