Reliable Beam Tracking on High-Altitude Platform for Millimetre Wave High-Speed Railway

High-speed railway (HSR) systems require reliable and high-capacity wireless communication to support passenger services, operational control, and safety-critical applications. However, maintaining uninterrupted connectivity for trains operating at speeds exceeding 300 km/h remains challenging, particularly for millimeter-wave (mmWave) communication due to severe path loss, signal blockage, and frequent beam misalignment. Conventional ground-based communication infrastructures often fail to provide consistent performance in such highly dynamic environments. This paper presents an integrated intelligent framework titled “Reliable Beam Tracking on High-Altitude Platform for Millimetre Wave High-Speed Railway”, which leverages High-Altitude Platforms (HAPs) and data-driven prediction techniques to enhance communication reliability. A supervised machine learning model is developed to predict mmWave beam reliability using key railway operational and geographical parameters such as train speed, population density, track information, and location attributes. In parallel, a deep learning-based Convolutional Neural Network (CNN) is employed to automatically detect structural defects in railway beams from image data, reducing reliance on manual inspections. Experimental results demonstrate that the proposed beam reliability prediction model achieves high classification accuracy, while the CNN-based defect detection model effectively identifies defective railway structures with strong confidence. The integration of both modules into a unified Flask-based web platform enables real-time prediction, automated infrastructure monitoring, and interactive data visualization. The proposed system offers a practical and scalable solution for improving communication stability and infrastructure safety in next-generation high-speed railway networks.