Transformer-Based Beam Prediction for THz Drone Communication with RIS Integration

This study addresses beam prediction in Terahertz (THz) Unmanned Aerial Vehicle (UAV) communication, a critical component of sixth-generation (6G) wireless networks. While the THz spectrum enables high-speed data transmission, UAV mobility introduces challenges like signal impairments and blockages. To overcome this, a deep learning model using self-attention networks is proposed, integrating Reconfigurable Intelligent Surfaces (RIS) to enhance coverage and adapt to channel dynamics. The approach predicts optimal beams for UAVs based on sequential data, including beam trajectories, position, and Line of Sight (LoS) information. Compared to Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRUs), the model demonstrates superior performance in prediction accuracy, as validated by metrics like Root-Mean-Square Error (RMSE) and Mean Absolute Error (MAE).