Terahertz (THz) communications is an emerging technology for sixth generation (6G) wireless communication systems and beyond with many challenges. Reaching Terabits per second (Tpbs) channel capacity in THz-band is physically crucial to the number of antennas at the base station (BS) and reconfigurable intelligent surface (RIS) elements, and requires careful innovative planning to manipulate electromagnetic (EM) waves. Throughout this letter, we propose joint beamforming (BF) and reflecting optimization for maximizing downlink (DL) data rate in THz-band communications equipped with RIS-aided orthogonal frequency division multiple access (OFDMA) and ultra-massive multiple-input multiple-output (UM-MIMO) system. Based on UM-MIMO channel model we derive, a non-convex problem is formulated in order to maximize the channel DL data rate. Then, we propose an alternating iterative algorithm to jointly optimize the BS BF weights, and RIS phase shift coefficients. In particular, the BS BF is solved by the gradient descent (GD) algorithm, while the RIS phase shift coefficients are solved by successive convex approximation (SCA) algorithm. Monte-Carlo simulation results are carried out to verify the efficiency and performance of the proposed framework and reveal that the DL data rate is significantly can be maximized.
