Distributed Intelligence Analysis Architecture for 6G Core Network

To achieve automation and intelligence in 5G networks, the 3rd Generation Partnership Project (3GPP) introduced the Network Data Analysis Function (NWDAF) as a novel network function. However, in the traditional 5G core network architecture, NWDAF relies on fixed configurations for data collection, lacking support for user customization and flexibility. Additionally, the current deployment of NWDAF is predominantly centralized, failing to provide real-time and reliable analysis services for the massive data in future 6G systems. Moreover, it is incapable of ensuring user privacy, making it incompatible with emerging scenarios like federated learning in 6G. Therefore, this paper proposes a user-customizable data collection approach and introduces a distributed NWDAF deployment based on the Raft algorithm, where the master node assigns data collection, analysis, and inference tasks to multiple worker NWDAFs. Our work and experimental results demonstrate that the proposed architecture effectively addresses these challenges and further achieves closed-loop network automation in 6G systems.