Location-Based Passive Beamforming for Rotatable RIS-Assisted Near-Field Communications

Exploiting the reconfigurable intelligent surface (RIS) to mitigate the severe path loss for millimeter-wave (mmWave) communications is a promising approach to meet the demand of ultra-high rates for future wireless communications, benefiting from its energy-efficient and cost-effective features. Especially, the near-field gain can be attained to improve the communication performance with its expanded aperture. However, achieving such superiority necessitates the acquisition of full channel state information (CSI) of RIS-associated channels, which poses a significant challenge in the near-field region, hindering the deployment of RIS. To solve this problem, in this paper we propose a location based scheme as an alternative to mitigate the large overhead of channel estimation for rotatable RIS-assisted near-field communications. In particular, we leverage the properties of Fresnel zone to construct the parametric equations of the Fresnel ellipsoids to characterize the propagation geometry. Then, in the presence of inevitable location uncertainty, we give the criterion of practical discrete phase shift design based on the derived parametric curve, resembling the minimum distance criterion for signal detection. Also, the rotation of RIS is exploited to mitigate the impact of location error. Such processing can further improve the performance of the proposed scheme, verified by analytical and numerical results.