Optimizing the Placement and Beamforming of RIS in Cellular Networks: A System-Level Modeling Perspective

TL;DR

This paper models the placement and beamforming of reconfigurable intelligent surfaces in cellular networks, demonstrating optimal placement strategies and proposing a novel beamforming method that significantly enhances system performance.

Contribution

It introduces a detailed 3D system-level model for RIS-assisted cellular systems, deriving optimal placement and a new beamforming design with improved sum rate performance.

Findings

Optimal RIS placement is directly opposite the base station for maximum coverage.

The proposed beamforming design outperforms existing algorithms in sum rate.

Significant performance gains are achieved with the new system-level modeling approach.

Abstract

In this letter, we present in detail the system-level modeling of reconfigurable intelligent surface (RIS)-assisted cellular systems by considering a 3-dimensional channel model between base station, RIS, and user. We prove that the optimal placement of RIS to achieve wider coverage is exactly opposite to the base station, under the constraint of single RIS in each sector. We propose a novel beamforming design for RIS-assisted cellular systems and derive the achievable sum rate in the presence of ideal, discrete, and random phase shifters at RIS. Through extensive system-level evaluations, we then show that the proposed beamforming design achieves significant improvements as compared to the state-of-the-art algorithms.