Robust Energy-Efficient Design for MISO Non-Orthogonal Multiple Access Systems
Robust Energy-Efficient Design for MISO Non-Orthogonal Multiple Access Systems
复制标题
DOI：
10.1109/tcomm.2019.2931972
复制
发表时间：
2019-07
影响因子：
8.3
通讯作者：
Faezeh Alavi;K. Cumanan;Milad Fozooni;Z. Ding;S. Lambotharan;O. Dobre
中科院分区：
文献类型：
--
作者：
Faezeh Alavi;K. Cumanan;Milad Fozooni;Z. Ding;S. Lambotharan;O. Dobre
非正交多址接入（NOMA）由于其频谱效率的显著增强而被设想为用于5G及以后无线网络的有前景的多址接入技术。在本文中，我们研究了多用户多输入单输出（MISO）NOMA系统的鲁棒能量效率设计，其中不完美的信道状态信息是在基站（BS）。该算法首先采用分簇算法将用户划分为不同的簇，然后采用NOMA技术在簇内用户之间公平地共享可用资源。为了消除集群之间的干扰，两种不同类型的迫零（ZF）的设计，即，混合ZF和全ZF，在BS。全ZF方案以更多的天线数量为代价完全消除了干扰泄漏，而混合ZF方案部分减轻了干扰泄漏。为了解决这个问题，Dinkelbach的算法被用来转换成一个简单的减法形式的非线性分式规划问题。最后，仿真结果表明，混合ZF优于全ZF计划与少数簇，而全ZF表现出更好的性能与更大的簇数。数值结果证实，我们提出的鲁棒方案优于非鲁棒方案在每个用户的速率满意率。
Non-orthogonal multiple access (NOMA) has been envisioned as a promising multiple access technique for 5G and beyond wireless networks due to its significant enhancement of spectral efficiency. In this paper, we investigate a robust energy efficiency design for multi-user multiple-input single-output (MISO) NOMA systems, where the imperfect channel state information is available at the base station (BS). A clustering algorithm is applied to group the users into different clusters, and then, the NOMA technique is employed to share the available resources fairly among the users in each cluster. To remove the interference between clusters, two different types of zero-forcing (ZF) designs, namely, hybrid-ZF and full-ZF, are employed at the BS. The full-ZF scheme completely removes the interference leakage at the cost of more number of antennas, and the hybrid-ZF scheme partially mitigates the interference leakage. To solve the problem, Dinkelbach’s algorithm is employed to convert the non-linear fractional programming problem into a simple subtractive form. Finally, simulation results reveal that hybrid-ZF outperforms the full-ZF scheme with a few clusters, while full-ZF shows a better performance with higher number of clusters. Numerical results confirm that our proposed robust scheme outperforms the non-robust scheme in terms of the rate-satisfaction ratio at each user.