IRS-Aided Near-Field Communication: Prospects and Challenges with Codebook Approach

Intelligent reflective surfaces (IRSs) offer a low-cost solution to reduced coverage in high-frequency bands used in next-generation communications by controlling the reflection of radio waves. However, as IRS size increases to enhance the propagation environment, the near-field region expands, leading to performance degradation in beamforming based on far-field assumptions. Beam focusing, which considers distance in addition to reflection direction, is more effective in the near field but results in overhead owing to beam training. To overcome this challenge, constructing a codebook that optimizes communication quality with fewer IRS control patterns is crucial. This study proposes a nonuniform three-dimensional (3D) codebook that optimizes IRS coverage by minimizing overlap and blank zones. The results confirm the effectiveness of the proposed codebook and show that its optimal shape depends on the desired control overhead. This article presents experimental results showing the effectiveness of beam focusing - a policy for constructing non-uniform 3D codebooks - and simulation evaluation results of communication performance when IRS is operated with various codebooks.