Trajectory-Based Handover Cell Selection Algorithm using GRU Model in 5G Networks

To achieve low latency and high-speed data transmission requirements, 5G networks support mmWaves and have implemented Heterogeneous Networks (HetNets). However, these deployments imply that 5G Base Stations (BSs) have a smaller coverage area, increasing the likelihood of Handover (HO) procedures and mobility issues. This article addresses the importance of selecting the appropriate cell for a handover situation, highlighting the importance of the user's position in managing fifth-generation (5G) networks. Also, it provides an algorithm that relates the user trajectory and Reference Signal Receive Power (RSRP) values to perform cell selection and handover decisions, employing a neural network model for trajectory prediction. Simulation results show that considering the simulated scenario, the handover frequency in the proposed solution decreased by 28.11% compared with the baseline solution, and the trajectory prediction presented a Mean Absolute Error (MAE) of 0.849.