Harmonizer Framework: Enhancing Control Plane Load Balancer to Meet URLLC Needs in NGN

The next-generation networks (Beyond 5G and 6G) are witnessing a tremendous increase in Control Plane (CP) signaling due to Service Based Architecture. The 5G Core deploys a Cloud Native Load Balancer (CNLB) to load balance CP signals between Cloud-native Network Functions (CNFs). Typically, 5G network slices like Ultra Reliable Low Latency Communications (URLLC) share the same CP resources across other slices. However, existing CNLB fails to balance CP resources between slice flows. Further, CNLB deploys simple scheduling (Weighted Round Robin and Least Connection) for load balancing and cannot address diverse needs of URLLC services like latency and reliability. We propose an extensible Harmonizer Framework that harmonizes diverse needs of different slice flows by leveraging AI/ML techniques. This enhances CNLB to prioritize and load balance CP signaling efficiently. Our simulation results show around 5-15 ms decrease in CP path setup time and 0.001-0.027% improved reliability of URLLC flows between User Equipments and CNFs.