A new Framework for Failure Classification in 5G Core Network

Considering the demands for high bandwidth, low latency, and high reliability, 5G network systems need continuous and high-quality operation to mitigate the consequences of service failures. To meet the diverse service requirements of 5G use cases, a cloud-based architecture based on NFV (Network Function Virtualization) and SDN (Software Defined Networks) technologies has been implemented. This has increased the complexity, workload and uncertainty of the operation and maintenance environment. Historically, fault management and other associated maintenance activities have been handled manually by humans. However, in the 5G era, relying entirely on humans will not be enough to cope with more demanding and complex network operation activities. A more automated approach to quickly analyze millions of network data to detect, classify or predict failures is thus needed. In this paper, we propose a ML-based system for autonomous 5G Core Network failure detection and classification. In addition, a comparative analysis of different applied machine learning/deep learning algorithms is presented.