A Handover Scheme Based on Mobility Prediction for Autonomous Moving Platforms in 5G Networks

In recent years, Autonomous Moving Platforms (AMP) have gradually become an essential part of industry and life, including a variety of water, land and air mobile platforms. 5G networks provide AMP with low-latency and high-speed communication services which can promote the construction of smart cities. However, the mobility of smart-AMP affects the handover in 5G networks. With the intensive deployment of gNBs, unreasonable handover will increase the power con-sumption of AMP and reduce the quality of service (QoS) of AMP communications. Thus, we propose a 5G handover method based on AMP trajectory prediction in this paper. A CNN-STLSTM neural network which combines a convolutional neural network (CNN) and Spatioal-Timporallong short-term memory (ST- LSTM) is proposed to predict the trajectory of AMP. Then, the gNB is selected to handover through the AMP's trajectory points predicted by the prediction model. The evaluation results show that the proposed mechanism can effectively reduce the number of hand over and improve the efficiency and stability of AMP networks.