Mobility-Aware Near-Field Beam Training in XL-MIMO Systems for High-Speed Railway

Near-field beam training is beneficial to avoid intricate direct channel estimation for extremely large-scale multiple-input-multiple-output (XL-MIMO) high-speed railway (HSR) systems. In this paper, we propose a mobility-aware near-field beam training (MNBT) scheme for the XL-MIMO system with orthogonal frequency-division multiplexing (OFDM) based on the periodicity and regularity of the train trajectory. The MNBT scheme can predict the beam direction and channel amplitude by utilizing the beam parameters obtained from the near-field graded beam training a few times initially, incorporating prior information of the HSR scenario such as the deterministic trajectory, motion model, and channel model. Based on the train’s speed, we develop an inter-carrier interference (ICI) suppression algorithm to mitigate the impact of Doppler frequency offset (DFO) induced by high-speed movement. Furthermore, to prevent sudden degradation of beamforming performance due to error accumulation, we establish an error detection module that identifies and corrects prediction errors in real time based on the concavity of the beam focusing function concerning the time variable. Simulation results demonstrate that, compared to existing beam training schemes, the proposed MNBT scheme significantly reduces beam training overhead while maintaining competitive beamforming performance in HSR scenarios.