This paper proposes a low-complexity and energy efficient sub-connected structure (SCS) hybrid beamforming (HBF) and fully-connected structure (FCS) HBF for non-orthogonal multiple access (NOMA) systems operating under line of sight (LOS) and non-line of sight (NLOS) millimeter wave (mmW) environments. The mathematical model to maximize sum-rate and energy efficiency (EE) of these two HBF-NOMA configurations is formulated to corroborate our simulation results. We aim to maximize both the sum-rates and EEs of the SCS-HBF-NOMA and FCSHBF-NOMA systems via successive interference cancellation-zero forcing (SIC-ZF) and phased-zero forcing (P-ZF) schemes, respectively. Associated performance analysis under New York University (NYU) mmW channel model is investigated for both two and four users per cluster. It is found that our proposed SCS-HBF-NOMA generally yields higher EE than the FCS-HBF-NOMA in both LOS and NLOS links having high signal to noise ratio (SNR) regime. Moreover, the proposed multistream HBF-NOMA leverages spatial multiplexing to attain higher sum-rates than the single stream counterpart. The employment of 3 bit quantization is capable of optimizing the sum-rate for energy efficient finite resolution HBFNOMA systems. Results also manifest that SCS-HBF structure with reduced antennas in its sub-array beamforms a wider beam capable of mitigating against the performance degradation on the account of channel imperfection in NOMA system. Furthermore, it is found that the number of users per cluster should be carefully selected for optimal performance of HBF-NOMA system. Lastly, space complexity of the HBF-NOMA scheme is found to be the lowest in P-ZF FCS-HBF-NOMA.

中文翻译：

---

多用户毫米波 NOMA 系统混合波束成形预编码器的性能分析


本文提出了一种低复杂度和高能效的子连接结构（SCS）混合波束成形（HBF）和全连接结构（FCS）HBF，用于在视距（LOS）和非视距 (NLOS) 毫米波 (mmW) 环境。制定了最大化这两种 HBF-NOMA 配置的总速率和能源效率 (EE) 的数学模型，以证实我们的模拟结果。我们的目标是分别通过连续干扰消除迫零（SIC-ZF）和相控迫零（P-ZF）方案来最大化 SCS-HBF-NOMA 和 FCSHBF-NOMA 系统的总速率和 EE。针对每个集群的两个和四个用户，研究了纽约大学 (NYU) 毫米波信道模型下的相关性能分析。结果发现，在具有高信噪比 (SNR) 状态的 LOS 和 NLOS 链路中，我们提出的 SCS-HBF-NOMA 通常比 FCS-HBF-NOMA 产生更高的 EE。此外，所提出的多流 HBF-NOMA 利用空间复用来获得比单流对应物更高的总速率。 3 位量化的采用能够优化节能有限分辨率 HBFNOMA 系统的总速率。结果还表明，子阵列波束中天线减少的 SCS-HBF 结构形成更宽的波束，能够减轻 NOMA 系统中由于信道缺陷而导致的性能下降。此外，我们发现每个集群的用户数量应该仔细选择，以获得 HBF-NOMA 系统的最佳性能。最后，发现 P-ZF FCS-HBF-NOMA 中 HBF-NOMA 方案的空间复杂度最低。