Optimization of Handover through Knowledge Graph Enabled Coordination between LLM and O-RAN

The evolution of telecommunication networks into 6G has introduced a paradigm shift, aiming to incorporate diverse and heterogeneous technologies along with targeting specialized use-cases tailored to private networks. Furthermore, the increased disaggregation of the network has enabled the ubiquity of Virtualized Network Functions (VNFs) to function on Commercially-Off-The-Shelf (COTS) hardware, opening an ecosystem of open interfaces and open software. Owing to this evolution, the management complexity of the network also grows correspondingly, mandating a shift in approach. In contrast to the static Policy Based Management and Control (PBMC) which requires the management system to be well thought out during the design phase, Artificial Intelligence (AI) is poised to learn and address the requirements during the run-time, introducing flexibility and adaptability, making it easier for the operator to manage and at the same time reducing operations and costs to design the management system. This article presents the incorporation of Large Language Model (LLM) adapted to Knowledge Graph (KG) for the purpose of intelligently triggering and coordinating handover in an Open Radio Access Network (O-RAN) system, thus anchoring KG as the intermediate representation and digital twin of knowledge.