IRS-aided Near-field Communication: Prospects and Challenges with Codebook Approach
Ryuhei Hibi, Hiroaki Hashida, Yuichi Kawamoto, Nei Kato
TL;DR
The paper addresses the near-field challenges of IRS-aided communication at high frequencies, where plane-wave-based beamforming becomes ineffective as IRS size grows. It advances a codebook-based beamtraining approach and introduces nonuniform 3D codebooks guided by a plane-wave approximation error analysis and a composite control accuracy index, supported by experiments and simulations. Key findings show that beamfocusing outperforms beamforming in the near-field, and that nonuniform 3D codebooks achieve higher SNR with fewer codewords compared to 2D or uniform 3D codebooks, while 2D codebooks can still be viable under strict overhead constraints. The work offers scalable, practical IRS control strategies for next-generation wireless systems, particularly for large-scale, millimeter-wave and terahertz deployments, enabling efficient near-field operation and paving the way for multiuser and mobile extensions.
Abstract
Intelligent reflecting surfaces (IRSs) are gaining attention as a low-cost solution to the coverage reduction in high-frequency bands used in next-generation communications. IRSs achieve low costs by controlling only the reflection of radio waves. However, to improve further the propagation environment, larger IRS sizes are required owing to their inability to amplify and retransmit signals. As the IRS size increases, the near-field region expands, requiring beamfocusing instead of beamforming, which is extensively used in existing research. This results in considerable overhead for IRS control decisions. To address this, constructing a codebook that achieves high communication quality with fewer IRS control patterns is effective. This article presents experimental results demonstrating the effectiveness of beamfocusing, construction policy for nonuniform three-dimensional codebooks, and simulation evaluation results of communication performance when operating IRSs with various codebooks. We believe these insights will foster further value for IRSs in next-generation communications.
