NATIVE OBSERVABILITY FOR PRIVATE 5G: A REFERENCE ARCHITECTURE ON COMBINATION OF NWDAF, SMO, AND EDGE ANALYTICS

Abstract

The transition to private 5G networks introduces unprecedented complexity in performance monitoring and assurance, demanding an integrated, data-driven observability framework. This paper proposes a novel reference architecture for native observability in private 5G, moving beyond fragmented tools to a cohesive, closed-loop system. The architecture strategically combines three core components: the 3GPP-standardized Network Data Analytics Function (NWDAF) for core-network intelligence and subscriber-aware analytics; the Service Management and Orchestration (SMO) framework for RAN-centric control and automation; and decentralized Edge Analytics platforms for ultra-low latency, applicationspecific telemetry processing at the network periphery.We detail how these components interoperate through open interfaces to create a unified data fabric, collecting, correlating, and analyzing metrics from the core, RAN, and edge applications. The proposed architecture enables predictive assurance, automated troubleshooting, and intent-based optimization, providing operators with a holistic, realtime view essential for maintaining stringent servicelevel agreements and operational efficiency in dedicated enterprise environments