A Decentralized Collaborative Learning Approach in 5G+ Core Networks

The 5G+ networks are going through extensive overhauling at both the radio access network (RAN) and core network (CN) to truly support the envisioned 5G+ use cases with ever so increasingly stringent quality of service requirements. Although many academic and industrial advancements have been realized, there are still needs for innovative changes vital for 5G+ connectivity, applications, and services. Consequently, the 3GPP is ushering in the next paradigm shift for CNs. CNs, which consist of several network functions (NFs) working cooperatively to service a RAN, are expected to exploit artificial intelligence (AI) in the form of an all-seeing-eye AI engine known as the Network Data Analytics Function (NWDAF). Its purpose is to optimize and automate network operation and management. However, accurate analytics require extensive machine learning (ML) training and as such, in this article, we present a decentralized collaborative ML approach to facilitate said training. Accordingly, we present two variations of a decentralized NWDAF architecture to enhance NF data localization, improve security, reduce control overhead during model training, shorten the training time, and finally, enhance the accuracy of the trained models by virtue of local testing on real-time network data. The first architecture decentralizes the global model training by exploiting the unique learning domain of each local model for each network function instance in a federated learning-based scheme while the second partitions the actual training network between multiple entities for each network function in a split learning-based scheme.

Introduction

The envisioned fifth-generation and beyond (5G+) networks are expected to support highly advanced services, applications, and use cases, all of which are characterized by extremely stringent quality of service (QoS) requirements. However, provisioning for such envisioned use cases presents significant challenges for mobile network operators [1]. In this light, various research propositions have indeed come to fruition at the radio access network (RAN) and core network (CN) in order to better meet the strict QoS requirements of 5G+ services. Nevertheless, further innovations are needed for mobile networks as a whole to truly satisfy the end-to-end QoS requirements. While the 5G use cases are ambitiously envisioned as enhanced mobile broadband (eMBB), massive machine type communications (mMTC), and ultra-reliable low latency communications (uRLLC), the 6G use cases are presented as far more lofty as further enhanced mobile broadband (feMBB), extremely reliable and low latency communications (ERLLC), ultra-massive machine type communications (umMTC), extremely low-power communications (ELPC), long-distance and high-mobility communications (LDHMC), massive ultra-reliable and low latency communications (muRLLC), and mobile broadband reliable and low latency communications (MBRLLC) [2], [3], [4]. Based on their naming scheme alone, there is a strong implication of an ever-tightening of QoS requirements.