Neural Networks Architecture and Hyperparameter Exploration for Handover Simulation in 5G Network

This paper presents a machine learning based approach for optimizing handover in 5G networks. By using the capabilities of neural networks, the proposed method enables proactive decision-making for target cell selection in handover process. Through analysis of key neural networks parameters, such as hidden nodes and epoch number, the neural network model predicts the target cell for handover. The simulations demonstrated the download success rate for each handover (HO). The results validate the advantages of using neural networks for handover optimization in 5G networks, providing seamless connectivity and improved user experience.