Low-Resolution Limited-Feedback NOMA for mmWave Communications
Low-Resolution Limited-Feedback NOMA for mmWave Communications
复制标题
DOI：
10.1109/twc.2020.2993212
复制
发表时间：
2019-10
影响因子：
10.4
通讯作者：
Yavuz Yapıcı;Ismail Güvenç;H. Dai
中科院分区：
文献类型：
--
作者：
Yavuz Yapıcı;Ismail Güvenç;H. Dai
频谱高效的毫米波（mmWave）通信作为解决频谱紧张问题的一种可行方案，近年来引起了人们的广泛关注。在这项工作中，我们提出了一种新的非正交多址接入（NOMA）框架，它利用毫米波通信的定向传播特性，从而提高频谱效率，通过非正交信令。特别地，我们认为一比特量化的角度信息是描述毫米波频带中的用户设备（UE）的信道质量的有限但有效的反馈方案。然后，不仅使用作为经典方法的1比特距离反馈，而且使用1比特角度反馈来建立用于NOMA传输的UE对。因此，所提出的策略被称为两位NOMA。我们还提出了一种新的混合策略，称为组合NOMA，通过两比特NOMA没有UE对的情况下。每当通过任何NOMA策略没有UE对可用时，我们采用具有适当UE选择方案的单用户传输（SUT）。混合中断和率性能进行了深入分析，与各自的中断和率的表达式。数值结果验证了该策略的性能优于单比特NOMA方案，无论是角度或距离唯一的反馈，并有一个非常接近的中断和率性能的最佳全分辨率反馈。
The spectrum-efficient millimeter-wave (mmWave) communications has recently attracted much attention as a viable solution to spectrum crunch problem. In this work, we propose a novel non-orthogonal multiple access (NOMA) framework, which makes use of the directional propagation characteristics of mmWave communications so as to improve the spectral efficiency through non-orthogonal signaling. In particular, we consider one-bit quantized angle information as a limited yet effective feedback scheme describing the channel quality of user equipment (UE) in mmWave bands. The UE pairs for NOMA transmission are then established using not only the one-bit distance feedback as a classical approach, but also the one-bit angle feedback. The proposed strategy is therefore referred to as two-bit NOMA. We also propose a novel hybrid strategy, called combined NOMA, for the circumstances with no UE pair through two-bit NOMA. Whenever no UE pair is available through any NOMA strategy, we resort to single user transmission (SUT) with proper UE selection schemes. The hybrid outage sum-rate performance is also analyzed thoroughly with the respective outage and rate expressions. The numerical results verify that the proposed strategy outperforms one-bit NOMA schemes with either angle- or distance-only feedback, and has a very close outage sum-rate performance to that for the optimal full-resolution feedback.