通信学报 ›› 2022, Vol. 43 ›› Issue (6): 179-188.doi: 10.11959/j.issn.1000-436x.2022120

• 学术论文 • 上一篇    下一篇

基于时延线阵列的毫米波NOMA系统混合预编码设计和功率分配

孙钢灿1,2, 吴新李1, 郝万明1,2, 朱政宇1,3   

  1. 1 郑州大学信息工程学院，河南 郑州 450001
    2 郑州大学产业技术研究院，河南 郑州 450003
    3 郑州大学电子材料与系统国际联合研究中心，河南 郑州 450001
  • 修回日期:2022-02-07 出版日期:2022-06-01 发布日期:2022-06-01
  • 作者简介:孙钢灿（1977- ），男，河南濮阳人，博士，郑州大学教授，主要研究方向为深度学习、机器学习、无线通信、物理层安全技术等
    吴新李（1995- ），男，河南固始人，郑州大学硕士生，主要研究方向为毫米波通信、NOMA无线通信、资源分配等
    郝万明（1988- ），男，河南安阳人，博士，郑州大学副研究员，主要研究方向为太赫兹通信、智能反射表面技术、边缘计算等
    朱政宇（1988- ），男，河南周口人，博士，郑州大学副教授，主要研究方向为无线通信与信号处理、智能反射表面技术、物理层安全技术等
  • 基金资助:
    国家自然科学基金资助项目(62101499);中国博士后科学基金资助项目(2020M682345);河南省博士后经费基金资助项目(202001015)

Hybrid precoding and power allocation for mmWave NOMA systems based on time delay line arrays

Gangcan SUN1,2, Xinli WU1, Wanming HAO1,2, Zhengyu ZHU1,3   

  1. 1 School of Information Engineering, Zhengzhou University, Zhengzhou 450001, China
    2 Industrial Technology Research Institute, Zhengzhou University, Zhengzhou 450003, China
    3 National Center for International Joint Research of Electronic Materials and Systems, Zhengzhou University, Zhengzhou 450001, China
  • Revised:2022-02-07 Online:2022-06-01 Published:2022-06-01
  • Supported by:
    The National Natural Science Foundation of China(62101499);China Postdoctoral Science Foundation Project(2020M682345);Henan Postdoctoral Foundation(202001015)

在线阅读

19

PDF下载

390

摘要：

目的：传统毫米波非正交多址系统中天线结构是基于高分辨率高能耗的移相器调制网络，如何降低系统能量消耗同时提高分辨率是需要解决的关键问题之一。本文将由开关和时延线组成的、可进行连续相位调制的低复杂度、低功耗的时延线阵列引入到毫米波非正交多址接入（NOMA）系统，研究了系统的能效和谱效。

方法：为降低用户间的干扰，提出改进K-means算法对用户进行分组，并为每组用户选择一个簇头，该算法将用户信道作为依据尽可能最大化同一簇用户信道相关性，降低不同簇用户间的相关性；然后根据簇头集合组成的相关用户信道矩阵，设计了一种低复杂度模拟预编码以最大化发射天线阵列增益，之后采用迫零技术设计数字预编码以消除波束间等效信道增益最大的用户间干扰；最后形成一个在用户服务质量和系统发射总功率约束下优化发送功率的能效最大化问题，针对所形成的非凸优化问题，提出一种两层迭代算法。具体来说，在外层应用Dinkelbach方法将能效优化中目标函数的分式结构转化为相减结构，在内层利用数学工具将非凸的目标函数转化为凸函数，然后提出一种基于交替优化（AO,alternating optimization）的迭代算法进行功率分配求解，最后通过内外两层循环迭代获得最初问题的解。

结果：仿真分析：（1）从图6(a)可以发现，所提方案下全连接和子连接混合预编码结构的系统谱效均高于传统基于移相器调制网络下全连接和子连接混合预编码结构的系统谱效；此外，在图6(a)中还可以看出，所提方案下全连接结构的系统频谱效率优于混合连接和子连接结构。（2）从图6(b)可以发现，所提方案下系统的能效也均优于传统基于移相器网络和全数字预编码结构的系统能效；同时，在图6(b)中还可以看出，所提方案下子连接结构的系统能效优于混合连接和全连接结构。（3）从图7(a))中可以发现，系统的谱效随着天线数的增加而增加；而从图7(b)可以发现，系统的能效随着天线数的增加而降低。（4）从图8(a)和(b)中可以发现，与基于时延线阵列的毫米波正交多址接入（OMA）系统相比，所提NOMA方案在能效和谱效方面均可以获得更优的性能。（5）所提出的改进K-means用户分组算法性能是优于K-means用户分组算法。K-means算法虽然可以根据信道的状态信息较好地完成用户分组，但其初始簇头的随机性将会影响算法收敛性和系统性能。另外，随机分组算法的性能最差，这是因为NOMA系统存在用户干扰，而随机分组将会使得同一簇内的用户干扰增大。

结论：时延线阵列是由的低功耗、低复杂度的开关和延迟线组成，通过调节开关可以实现连续相位调制。与基于时延线阵列的毫米波OMA系统和传统基于移相器的毫米波NOMA系统相比，本文构建的基于时延线阵列的毫米波NOMA系统可以有效的提高系统的谱效和能效。

关键词: 毫米波, 非正交多址接入, 混合预编码, 功率分配

Abstract:

Objectives: The antenna structure in conventional millimeter-wave (mmWave) nonorthogonal multiple access(NOMA)systems is based on a high-resolution and high-energy phase shifter modulation network,and how to reduce the system energy consumption while improving the resolution is one of the key problems to be solved. In this thesis, a lowcomplexity and low-power delay line array consisting of switches and delay lines for continuous phase modulation is introduced into the mmWave NOMA system, and the energy efficiency(EE)and spectral efficiency(SE)of the system are investigated.

Methods:To reduce inter-user interference,an improved K-means algorithm is proposed to group users and select a cluster head for each group of users, which maximizes the correlation of user channels in the same cluster and reduces the correlation between users in different clusters as much as possible;Then,a low-complexity analog precoding is designed to maximize the array gain of the antenna based on the relevant user channel matrix composed of the cluster head set, followed by a digital precoding designed to eliminate inter-user interference with the maximum equivalent channel gain between beams using a forced-zero technique; Finally, an EE maximization problem is formed to optimize the transmit power under users'quality of services and total transmit power constraints,and a two-layer iterative algorithm is proposed for the resulting non-convex optimization problem. Specifically,the Dinkelbach method is applied in the outer layer to transform the fractional structure of the objective function in EE optimization into a subtractive structure, and the non-convex objective function is transformed into a convex function by using mathematical tools in the inner layer, and then an iterative algorithm based on alternating optimization (AO) is proposed for power allocation, and finally the solution of the initial problem is obtained by circular iteration in both inner and outer layers.

Results:Simulation analysis:(1)The figure6(a)show that the SE of the proposed scheme is higher than that of the conventional phase-shifter modulation network with fully-connected and sub-connected hybrid precoding structures; In addition, the system SE of the fullconnected structure under the proposed scheme is better than that of the hybrid-connected and sub-connected structures.(2) the figure6(b) show that the EE of the system under the proposed scheme is also better than that of the conventional phase-shifter network-based and fully-digital precoding structures; Also, the system SE of the sub-connected structure under the proposed scheme is better than that of the hybrid-connected and fully-connected structures. (3) The figure7(a) show that the SE of the system increases as the number of antennas increases;The figure7(b)show that the EE of the system decreases as the number of antennas increases. (4) The figure8(a) and (b) show that compared with the mmWave orthogonal multiple access (OMA) system based on the time-delayed line array, and the proposed scheme can achieve better performance in terms of EE and SE.(5)the figure9(a) show that the performance of the proposed improved K-means user grouping algorithm is better than the K-means user grouping algorithm.Although the K-means can be done better based on the state information of the channel,the randomness of its initial cluster head will affect the convergence of the algorithm and the system performance. In addition, the random grouping algorithm has the worst performance because there is user interference in NOMA system,and the random grouping will make the user interference in the same cluster increase.

Conclusions: The time-delay line array is composed of low-power, low-complexity switches and delay lines,and the continuous phase modulation can be realized by adjusting the switches.The proposed mmWave NOMA system based on delay line array constructed can effectively improve the SE and EE of the system.

Key words: mmWave, NOMA, hybrid precoding, power allocation

中图分类号: 

No Suggested Reading articles found!