O-RAN based proactive ANR optimization

As telecommunication networks evolve towards 5G, emerging technologies such as Massive Internet of Thing (mIoT) and massive Machine Type Communication(mMTC) are predicted to add billions of wireless devices to the 5G network. Tightly coupled Radio Access Network (RAN) devices are one of the major bottlenecks in the expansion of 5G network. To cater the dynamic and massive demand of the network,requires decoupling the network device hardware with its software function. Open Radio Access Network (ORAN) framework aims to achieve the decoupling of RAN device hardware with its function software to achieve auto-scaling of RAN network functions to meet the ever increasing and dynamic demand for network access. The 5G network needs to be densified to increase the capacity and coverage of networks. However network densification comes with its own challenges, as number of cells increase, it becomes more complex to manage and optimize neighbour relationships. Automatic Neighbour Relation (ANR) is a well know Self Organizing Network (SON) function that is used to manage neighbour cell relationships,optimization of the Neighbour Cell Relation Table (NCRT), significantly improves the handover timing, reduces the call drop rates and increase the total number of successful handovers. This paper investigates a new approach for ANR optimization for the next generation networks using O-RAN defined open interfaces and architectural platform. The approach would leverage O-RAN architecture that supports implementation of intelligent models and proposes a Machine Learning (ML) based proactive ANR optimization technique to improve gNodeB (gNB) handovers.