Pre-Emptive Conflict Detection Architecture for O-RAN Service Management and Orchestration

The evolution of telecommunications networks from closed to open architectures has led to the development of the Open Radio Access Network (O-RAN) standard. O-RAN's Service Management and Orchestration (SMO) typically contains multiple autonomic management agents that simultaneously make configuration changes to optimize aspects of the network. Dealing with competing and conflicting configuration changes from these autonomic management agents remains a major issue. This paper proposes an extension to the O-RAN SMO architecture to incorporate a component that can predict when configuration changes are likely to degrade network performance. The purpose of this approach is to safeguard the actions taken by the management agents against unintended effects. The approach involves the use of two levels of clustering based on cell similarity and cell configuration. The results consist of predictions of network performance degradation based on statistical correlation with network performance.