On the Performance of IRS-Assisted Multi-User THz-NOMA Under Beam Misalignment

This paper considers an intelligent reflecting surface (IRS)-assisted multiple-input single-output multi-user non-orthogonal multiple access (NOMA) network operating in the terahertz band subject to co-channel interference (CCI). The IRS is examined under various configurations: i) IRS partitioning with discrete phase shifts, ii) IRS with random phase shifts, and iii) IRS with discrete phase shifts for one user and random phase shifts for the remaining users. It aims to assess the impacts of various factors, including beam misalignment, imperfect successive interference cancellation (SIC), CCI, phase quantization errors, and random phase shifts. Thus, it derives accurate expressions for the ergodic rate, outage probability and diversity order, which are verified via numerical and simulation results. The results show that the impacts of beam misalignment and imperfect SIC are more dominant compared to those resulting from random phase shifts, phase quantization errors and CCI. Despite achieving lower performance compared to the discrete phase shifting scheme, the random phase shifting scheme can be a promising candidate for multi-user NOMA, as it avoids the costly channel estimation and computational complexity that are proportional to the numbers of IRS elements and users. The results also reveal that partitioning the IRS under discrete phase shifts outperforms a combination of discrete and random phase shifts. The system performance in the high-signal-to-noise ratio regime is also analyzed, which shows that the performance reaches a limit due to inter-user interference and CCI.