Exploiting Beam Split-Based Multi-User Diversity in Terahertz MIMO-OFDM Systems

The wide availability of TeraHertz (THz) bands, in conjunction with hybrid beamforming based Multi-Input Multi-Output (MIMO) and Orthogonal Frequency Division Multiplexing (OFDM) technologies, promises unprecedented capabilities for future 6G networks. However, these systems are degraded by the beam split effect, a THz-specific issue where the frequency-independent phase shifters of hybrid beamforming lead to the deviation of the intended beam depending on subcarrier frequencies. Instead of mitigating this effect as in usual approaches, we propose to exploit it as an opportunity to enhance global multi-user THz MIMO-OFDM performances, through joint user grouping and beam split multiplexing across users for optimizing key multi-user objectives, giving rise to beam split-based multi-user diversity. We design two beam split-based resource allocation methods where user assignment and subcarrier allocation are optimized, firstly to maximize the overall array gain over subcarriers, and secondly, to maximize system spectral efficiency. Numerical results show that our beam split-tailored methods enable to improve the performance of dense multi-user THz systems as compared to benchmarks.