Sensing-Aided High-Efficiency Hybrid Precoding for Wideband Terahertz Massive MIMO Systems

The beam split in the frequency domain seriously affects hybrid precoding performance in wideband terahertz (THz) massive multiple-input multiple-output (MIMO) systems. The receiver hardly catches full Channel State Information (CSI) of all subcarriers due to spatial angle diffusion caused by the beam split, which would decrease the efficiency of hybrid precoding. To address the decrease, sensing-aided hybrid precoding is proposed to deduce the frequency-angle information in this paper. A channel matrix inverting scheme which can regain full CSI from the frequency-angle information of partial CSI is modeled. A 300 GHz MIMO system with 15 GHz bandwidth is simulated to validate the proposed sensing-aided hybrid precoding. Spectral and energy efficiency are employed to characterize the performance of the systems. Compared with true-time-delayers (TTD) based and partial CSI based schemes with no sensing in the state of the art, the proposed sensing-aided scheme performs higher spectral and energy efficiency.