Reconfigurable Intelligent Surface Based Hybrid Precoding for THz Communications
share
Terahertz (THz) communication has been considered as a promising technology to provide ultra-high-speed rates for future 6G wireless systems. To alleviate the severe propagation attenuation in THz communication systems, massive multiple-input multiple-output (MIMO) with hybrid precoding can be used for beamforming to provide high array gains. In this paper, we propose a reconfigurable intelligent surface (RIS)-based hybrid precoding architecture for THz communication, where the energy-efficient RIS rather than the energy-hungry phased array is used to realize the analog beamforming of the hybrid precoding. Then, we investigate the hybrid precoding problem to maximize the sum-rate for the proposed RIS-based hybrid precoding architecture. Due to the non-convex constraint of discrete phase shifts by considering the practical hardware implementation of RIS, this sum-rate maximization problem is challenging to solve. Thus, we reformulate it as a parallel deep neural network (DNN)-based classification problem, which can be solved by the proposed deep learning-based multiple discrete classification (DL-MDC) hybrid precoding scheme with low complexity. Simulation results show that the proposed scheme works well both in theoretical Saleh-Valenzuela channel model and practical 3GPP channel model, and it can reduce the runtime significantly with a negligible performance loss compared with existing iterative search algorithms.
READ FULL TEXT
