Passive Beamforming Design for Reconfigurable Intelligent Surface-aided OFDM: A Fractional Programming Based Approach.

Keming Feng,Yijian Chen,Yu Han,Xiao Li,Shi Jin
DOI: https://doi.org/10.1109/vtc2021-spring51267.2021.9449003
2021-01-01
Abstract:Reconfigurable intelligent surface (RIS) is a low-cost device envisioned to achieve substantial promotion in both spectrum and energy efficiency in the future wireless communication systems. In this paper, we investigate the downlink transmission of the RIS-enabled orthogonal frequency division multiplexing (OFDM) system, and propose a low-complexity passive beamforming optimization algorithm to maximize the achievable sum-rate over all subcarriers. The passive beamforming optimization is a NP-hard problem due to the RIS-induced non-convex unit modulus constraints. We conquer this difficulty by exploiting a fractional programming (FP) based approach combined with an efficient manifold optimization (MO) method. With the proposed low-complexity passive beamforming optimization algorithms and water-filling power allocation, the achievable sum rate can then be maximized through alternating optimization (AO). Simulation results indicate that the proposed AO based algorithm performs well in achieving high average sum-rate with a fast convergence rate.
What problem does this paper attempt to address?