Mobile User Trajectory Prediction Based On Machine Learning

5G network brings users a smoother Internet experience,but in 5G ultradense network,the number of base stations is large,the coverage of a single base station is small,and the energy consumption of base stations is high,which will lead to frequent cell switching and increased operator costs.One of the key points to solve these two problems is the trajectory prediction of mobile users.Trajectory prediction and its application in 5G scenarios will help to improve user experience and promote energy saving of base stations.With the development of global positioning technology and the popularization of intelligent devices,it is possible to obtain the historical trajectory of mobile users anytime and anywhere.However,in recent years,the amount of trajectory data is increasing,and the accuracy of trajectory prediction is also required to be higher in various scenarios.Traditional trajectory prediction methods are facing increasing challenges.At the same time,neural network as an emerging technology is being applied to trajectory prediction,especially long short-term memory network,which is suitable for processing time series,will bring a new solution to the problem of trajectory prediction.This paper uses machine learning methods for mobile user trajectory prediction and applies it to 5G scenarios.The main work is as follows:1.Using Long Short-Term Memory model,a trajectory prediction method is proposed.Simulation results show that the error distance of the proposed method can be less than 3 meters in 1s.Compared with other methods,the proposed method performs better in terms of mean absolute error,mean square error and other indicators.2.The proposed trajectory prediction method is applied to resource reservation.Simulation results show that on the test data set used in this paper,the cell prediction accuracy based on trajectory prediction can reach more than 84%,and the resource reservation scheme can reduce the average waiting delay of users by more than 80%.3.The proposed trajectory prediction method is applied to 5G base station energy saving management.The simulation results show that on the test data set used in this paper,the cell prediction accuracy based on trajectory prediction can reach 94.155%,and the energy saving scheme of base station can reach 8.3%energy saving efficiency.