DEVELOPMENT OF BEAM TRACKING PREDICTION SYSTEM FOR 5G HIGH-SPEED TRAINS USING MACHINE LEARNING

• Main Author: Jonathan
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/82401
• Institution: Institut Teknologi Bandung

High-speed trains generally still use the GSM-R communication system, which is a 2G technology specifically designed for railway communication lines and is far behind compared to today's technology. One of the reasons why the railway communication system still uses GSM-R is because the data transmission speed is not the main thing, but rather the security of the data transmission. With the development of the 5G communication system, both of these things can be achieved and used in the railway communication system. However, this is difficult to implement because the 5G communication system uses mmWave which causes a small beam bandwidth. The small beam bandwidth is an obstacle because to get optimal data transmission, it requires good beam accuracy from the base station to the train, while the high-speed train is moving at high speed. This research is aimed at overcoming these difficulties. By implementing machine learning computing algorithms, a beam tracking prediction system will be created to improve the accuracy of the beam from the base station to the train. The machine learning algorithms analyzed in this study are K-Nearest Neighbors, Neural Network, Lookup Table, Random Forest, Support Vector Machine, and Naive Bayes. The research methods used include literature studies, system design, testing, and performance analysis. The results of the study show that the implementation of the Neural Network computing algorithm produces the best beam forming accuracy and speed.