Alternating Optimization Based Hybrid Beamforming in Terahertz Widely-spaced Multi-subarray Systems.

With multi-GHz bandwidth, Terahertz (THz) communication has drawn increasing attention in the sixth generation (6G) and beyond communication systems. Although different hybrid beamforming algorithms have been developed to improve array gain and spectral efficiency in both single-user and multiuser systems, they rely on the compact array architecture and far-field approximation of propagation. In the THz band, due to the limited number of multipath and shortage of spatial degree freedom of the channel, the spatial mulitplexing gain is constrained. In this paper, a widely-spaced multi-subarray (WSMS) hybrid beamforming architecture is proposed in a multi-user network, to harness both the inter and intra-multiplexing gains in the enlarged near-field propagation region. Specifically, the novel structure of the hybrid beamforming matrices is analyzed, and an alternating optimization algorithm to maximize the sum rate is designed. Simulation results demonstrate that the proposed algorithm under the WSMS architecture outperforms the existing beamforming schemes in the traditional architectures.