Informer Based Channel Prediction with Multiple Predictor Antennas for High-Speed Railway
Zhaoming Dai, Jiakang Zheng, Jiayi Zhang, Bo Ai · 2024
The use of predictor antennas (PAs) has significant potential to enhance wireless channel prediction performance in high-speed railway (HSR) communications. The PA system features two sets of antennas installed on the roof of a vehicle. The PA is located at the front of the vehicle and is used to predict the channel observed by the receive antenna (RA), which is located behind the PA. The PAs can be integrated with dense pilots spatially, but the prediction performance decreases when channel estimations are sparse. Therefore, this paper first proposes a multiple PAs (mPAs) system combined with interpolation for sparse channel estimations. Subsequently, recognizing the need for PAs to measure all antenna channels in the estimation interval, we propose an informer-based mPAs system. This system predicts future RA channels in parallel, effectively solving the problem of error propagation in sequential prediction methods. Simulation results demonstrate that as the prediction horizon extends, the proposed informer-based mPAs system outperforms others. Finally, we investigate how varying velocities impact prediction accuracy. It was found that the prediction accuracy of a single PA system performs well at low speeds but drops rapidly at high speeds. Moreover, our proposed informer-based mPAs system achieves higher prediction horizons and maintains efficiency at high speeds.