Real-Time Monitoring of 5G Networks: An NWDAF and ML Based KPI Prediction

The emergence of 5G networks plays a crucial role in satisfying the diverse demands of the rapidly evolving telecommunications market. The introduction of Network Function Virtualization (NFV) and Software-Defined Networks (SDN) offers significant networking advantages by transforming hardware-dependent network functions into programmable and cost-efficient Virtual Network Functions (VNFs). However, for effective monitoring strategies and enforcing Service Level Agreements (SLAs), these virtualized networks must be made visible (observable) to service providers or network operators. Extensive research has been undertaken to address monitoring and analytics challenges as networks transition to virtualized environments. This paper presents the development and evaluation of a real-time monitoring solution for 5G Service-Based Architecture (SBA) by leveraging a Network Data Analytics Function (NWDAF) developed by the authors. This solution allows and demonstrates real-time monitoring of 5G networks via a standardized interface. In addition, NWDAF is extended to utilize Machine Learning (ML) models to predict Key Performance Indicators (KPIs) related to network traffic, enhancing the overall network analytics capabilities. Furthermore, the collected metrics from the 5G core are pushed into a Prometheus database, with Grafana providing a visual analytics tool for assessing network performance.