5GC-Analyser: Demistifying the 5G Core Network Through Statistical Analysis

Observability has become a crucial aspect in cloud native environments, allowing for the measurement of internal states in IT systems. With the emergence of 5G and the anticipation of 6G technologies, the significance of observability in mobile networks has been further emphasized. This paper presents, 5GC-Analyser, a novel approach to understanding the functionality of the 5G core network (5GC) through statistical analyses. By exploring the statistical relationships and causalities among various 5GC microservices and their metrics, our 5GC-Analyser solution, enables the selection of appropriate techniques for online observability. Leveraging these analyses, our framework provides a comprehensive understanding of the intricate dynamics within the 5GC, facilitating improved network management and optimization strategies. We validate our solution in a real-world 5G scenario by simulating end users and replicating conditions observed in a real Orange's 5G gNodeBs (gNB). The obtained results pave the way for future work in proactive detection of performance degradation in the 5G core network. This research bridges the gap between theory and practice, offering valuable insights for network operators and researchers in the field of 5G technology.