Beam Misalignment Detection-Based Beam Training in Near-Field Communications

Extremely large-scale array (XL-array) is an emerging technology that is expected to meet the high data rate demands of the future 6G networks. However, as the number of antennas significantly increases, users are likely to be located in the region considered as the near-field. To perform beam training suited for near-field systems, a polar codebook that searches both angular and distance dimensions is used. An exhaustive search method utilizing the polar codebook exists, which explores all interest regions. However, due to the requirement of searching both angular and distance domains, this approach leads to a severe beam training overhead. Therefore, in this paper, we propose a novel hierarchical beam training algorithm that dramatically reduces the beam training overhead while ensuring reasonable performance. This algorithm significantly reduces the beam training overhead compared to the exhaustive search and, as shown in the numerical results, even outperforms it in certain regions.