Deep Learning-Based Intelligent Framework for Cloud-Native 5G Core Network Management and Optimization

Abstract

Cloud native technology, which offers previously unheard-of levels of operational robotics, scalability, and versatility, has completely transformed 5G and 6G communications networks. However, distributing resources for fluid cloud computing environments faces a new difficulty due to the wide range of cloud native services and apps. The proposed framework, based on the CygNet MaSoN architecture, integrates real-time monitoring, data aggregation, predictive analytics, and deep learning models to optimize resource utilization and detect anomalous network behavior. The system enables proactive identification of service degradation, network performance issues, and security threats while supporting self-organizing and closed-loop automation capabilities required for autonomous 5G networks. Furthermore, the framework incorporates sequence-aware learning algorithms and synthetic data generation techniques to improve model performance in dynamic and context-dependent network environments. The components of a system and architecture are described in detail. After that, three actual use cases that have been performed on this structure are explained. The features taken into consideration are discussed together with machine learning in general models created and synthetic data production techniques used. These findings support the significance of sequence-aware algorithms for protecting roaming environments, which frequently involve context-dependent and fleeting dangers. The suggested paradigm offers a route for robust security in networks beyond 5G and 6G as well as a basis for intelligent, adaptive security monitoring in 5G.