Dynamic Delay-Doppler-Angle Domain Channel Tracking for THz Massive MIMO-OTFS Communication Systems

As data transmission demand grows rapidly, terahertz (THz) frequency has been regarded as the most promising candidate in future sixth-generation (6G) communication systems. Facing severe path loss, THz communications have to use massive MIMO technology to promote the directional transmission gain in narrow beams. However, narrow beams and wide bandwidth signal in THz communications together result in the time-frequency-space non-stationary channel. To track this non-stationary channel, we first combine THz massive MIMO communication system with orthogonal time frequency space (OTFS) modulation to get the union sparse channel on delay-Doppler-angle (DDA) domain. Then this non-stationary channel is modeled through our proposed DDA-2D-Markov models (DDA-2D-MM) considering the unique channel characteristics on delay, Doppler and angle dimension respectively. Besides, we design a unique DDA-structured approximate message passing (DDA-STAMP) algorithm to track the dynamic channel with the prior information in consecutive time slots to improve the algorithm convergence speed and accuracy performance. Simulation results show that the proposed DDA-STAMP algorithm enhances both of the time consumption and accuracy performance in THz massive MIMO-OTFS communication channel tracking.