Position-Attitude Prediction based Beam Tracking for UAV mmWave Communications

Millimeter wave offers large bandwidth for high data-rate unmanned aerial vehicle (UAV)-to-UAV communications. Because of high mobility and attitude variations, it is challenging to maintain the communication link among the navigating UAVs with narrow beam in the mmWave band. To the best of our knowledge, this is the first paper to establish a transmission-oriented UAV attitude prediction model for the UAV-to-UAV mmWave communication link. In particular, a position-attitude prediction based beam tracking algorithm is proposed. First, a Guassian Process (GP) based learning algorithm is presented for the transmitting UAV to predict the position and attitude of the receiving UAV by using the previous position-attitude data and exploiting the relationship between the position and attitude. Then, the analog beamforming vectors are derived by using the predicted spatial angles. Simulation results demonstrate that the proposed learning algorithm can achieve high accurate position-attitude prediction, and the beam tracking algorithm considering UAV attitude variations significantly outperforms the existing algorithms with only position information.