Intelligent reflective surfaces (IRSs) are effective for extending the communication range of base stations whilst minimizing costs; however, they do not amplify radio waves. Therefore, the size of the reflective surface must increase to increase gain. In the far field, differences in phase rotation amount through each IRS element depend only on the direction. Therefore, control decisions for the IRS can be made through beam training, which involves searching for a 2D codebook. However, with an increase in the size of the IRS, the phase difference varies with distance in the enlarged near field, even in the same direction. Thus, the control decisions of the IRS require 3D information, which is the distance, angle, and a 3D codebook. In general, 3D codebooks are larger than 2D codebooks, thereby incurring a large control overhead for beam training. Thus, in this study, we propose a 3D codebook for the near-field extension of large IRSs. The proposed 3D codebook can flexibly balance the communication quality and overhead by modeling the accuracy required for IRS control at each location, enabling efficient IRS control in response to the communication requirements. The performance of the proposed method is evaluated using simulations.
3D Codebook Construction Strategy based on Control Accuracy Index for Intelligent Reflecting Surface in Near-field
07.10.2024
482093 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch