Joint time-angle domain sparse channel prediction for large-scale antenna array millimeter-wave communications

In large-scale antenna array millimeter-wave communication systems, the channel state information exhibits rapid variability in motion. Traditional channel estimation methods cannot react to the changes in real time. A time-angle domain joint sparse channel prediction method is proposed, which utilizes the signal characteristics of different representation domains of the large-scale antenna array channel. The method jointly analyzes the dual sparse characteristics of the channel's angle domain and time domain. By modeling the angular distribution of the channel, the sparse characteristics in the angle domain are effectively utilized. Through the analysis of the channel's time domain multipath information, the sparse characteristics of time domain delays are efficiently utilized. The classic orthogonal matching pursuit algorithm is employed for joint sparse reconstruction of the channel's angle and time domains, with iterative optimization during the channel prediction process. The simulation experiments construct millimeter-wave multipath channels based on the 3rd generation partnership project (3GPP) channel model, conducting comparative analyses on the proposed channel prediction method in terms of accuracy and performance metrics including precoding bit error rate (BER). Experimental results show that, compared with traditional single-domain channel prediction methods, the time-angle domain joint prediction method significantly improves the accuracy and stability of channel prediction. The research findings are of significant importance for adaptive transmission and beamforming in large-scale antenna array systems.