基于DNN的Sub-6 GHz辅助毫米波网络功率分配算法

doi:10.11959/j.issn.1000-436x.2021170

通信学报 ›› 2021, Vol. 42 ›› Issue (9): 184-193.doi: 10.11959/j.issn.1000-436x.2021170

基于DNN的Sub-6 GHz辅助毫米波网络功率分配算法

孙长印¹^,², 刘李延¹, 江帆¹^,², 姜静¹^,²

¹ 西安邮电大学通信与信息工程学院，陕西西安 710121
² 西安邮电大学信息通信网络与安全重点实验室，陕西西安 710121

修回日期:2021-07-01 出版日期:2021-09-25 发布日期:2021-09-01
作者简介:孙长印（1963− ），男，陕西扶风人，博士，西安邮电大学副教授、硕士生导师，主要研究方向为无线异构网络干扰管理、资源分配技术等
刘李延（1998− ），男，陕西咸阳人，西安邮电大学硕士生，主要研究方向为毫米波通信与功率控制技术
江帆（1982− ），女，江苏盐城人，博士，西安邮电大学教授，主要研究方向为无线资源管理、D2D 通信技术边缘计算及缓存技术等
姜静（1974− ），女，陕西安康人，博士，西安邮电大学教授，主要研究方向为人工智能在无线通信中的应用与Massive MIMO技术
基金资助:
国家自然科学基金资助项目(61801382);国家自然科学基金资助项目(61871321);国家自然科学基金资助项目(62071377);陕西省自然科学基金资助项目(2019JZ-06);陕西省重点产业链项目(2020ZDLGY02-06);陕西省重点产业链项目(2019ZDLGY07-06)

DNN-based Sub-6 GHz assisted millimeter wave network power allocation algorithm

Changyin SUN¹^,², Liyan LIU¹, Fan JIANG¹^,², Jing JIANG¹^,²

¹ School of Communication and Information Engineering, Xi’an University of Posts and Telecommunications, Xi’an 710121, China
² Key Laboratory of Information Communication Network and Security, Xi 'an University of Posts and Telecommunications, Xi’an 710121, China

Revised:2021-07-01 Online:2021-09-25 Published:2021-09-01
Supported by:
The National Natural Science Foundation of China(61801382);The National Natural Science Foundation of China(61871321);The National Natural Science Foundation of China(62071377);The Natural Science Foundation of Shaanxi Province(2019JZ-06);The Key Chain Project of Shaanxi Province(2020ZDLGY02-06);The Key Chain Project of Shaanxi Province(2019ZDLGY07-06)

摘要/Abstract

摘要：

针对毫米波系统功率控制测量伴随的信令开销与功耗，以及迭代操作带来的复杂度问题，提出了一种使用Sub-6 GHz频段预测毫米波链路功率分配的算法。首先，分析了Sub-6 GHz频段信道信息到毫米波频段最佳功率分配的映射关系。其次，基于这种映射关系设计了一种深度神经网络（DNN）模型，通过使用加权最小均方误差（WMMSE）准则对神经网络进行了不同场景的训练，实现利用Sub-6 GHz频段对毫米波频段信道的功率分配进行预测。仿真结果表明，所提算法在仅采用 Sub-6 GHz 频段信道信息的情况下，与毫米波频段下的 WMMSE算法相比，可在耗时少于其0.1%的同时，获得其不小于97%的和速率性能。

关键词: 毫米波通信, 功率分配, 深度学习, 神经网络

Abstract:

Aimed at the problems of the signaling cost and power consumption in the power control measurement of the millimeter wave system, as well as the complexity caused by iteration operations, a millimeter wave link power allocation prediction algorithm using the Sub-6 GHz frequency band was proposed.Firstly, the mapping between the Sub-6 GHz band channel information and the optimal power allocation of the millimeter wave band was analyzed.Then, a deep neural network (DNN) model was utilized to realize this mapping function.To predict the power allocation of millimeter wave channel with Sub-6 GHz channel as input, the neural network was trained with the weighted mean square error minimization method (WMMSE) as the supervisor in different scenarios.The simulation results show that compared with the WMMSE algorithm in millimeter wave band, the proposed algorithm can obtain more than 97% of its sum-rate performance while taking less than 0.1% of the time.

Key words: millimeter wave communication, power allocation, deep learning, neural network

中图分类号:

TN928

孙长印, 刘李延, 江帆, 姜静. 基于DNN的Sub-6 GHz辅助毫米波网络功率分配算法[J]. 通信学报, 2021, 42(9): 184-193.

Changyin SUN, Liyan LIU, Fan JIANG, Jing JIANG. DNN-based Sub-6 GHz assisted millimeter wave network power allocation algorithm[J]. Journal on Communications, 2021, 42(9): 184-193.

图/表 12

图1

表1

表2

图2

图3

表3

图4

图5

表4

表5

图6

图7

参考文献 26

[1]	RAPPAPORT T S , XING Y C , KANHERE O ,et al. Wireless communications and applications above 100 GHz:opportunities and challenges for 6G and beyond[J]. IEEE Access, 2019,7: 78729-78757.
[2]	YU Y W , DUTKIEWICZ E , HUANG X J ,et al. Downlink resource allocation for next generation wireless networks with inter-cell interference[J]. IEEE Transactions on Wireless Communications, 2013,12(4): 1783-1793.
[3]	SHI J F , CHEN M , YANG Z H ,et al. Power control and performance analysis for full-duplex relay-assisted D2D communication underlaying fifth generation cellular networks[J]. IET Communications, 2017,11(18): 2729-2734.
[4]	SHI Q J , RAZAVIYAYN M , LUO Z Q ,et al. An iteratively weighted MMSE approach to distributed sum-utility maximization for a MIMO interfering broadcast channel[J]. IEEE Transactions on Signal Processing, 2011,59(9): 4331-4340.
[5]	BALIGH H , HONG M Y , LIAO W C ,et al. Cross-layer provision of future cellular networks:a WMMSE-based approach[J]. IEEE Signal Processing Magazine, 2014,31(6): 56-68.
[6]	YU W , CIOFFI J M . FDMA capacity of Gaussian multiple-access channels with ISI[J]. IEEE Transactions on Communications, 2002,50(1): 102-111.
[7]	SUN H R , CHEN X Y , SHI Q J ,et al. Learning to optimize:training deep neural networks for wireless resource management[C]// Proceedings of 2017 IEEE 18th International Workshop on Signal Processing Advances in Wireless Communications (SPAWC). Piscataway:IEEE Press, 2017: 1-6.
[8]	RUSEK F , PERSSON D , LAU B K ,et al. Scaling up MIMO:opportunities and challenges with very large arrays[J]. IEEE Signal Processing Magazine, 2013,30(1): 40-60.
[9]	NGO H Q , LARSSON E G , MARZETTA T L . Energy and spectral efficiency of very large multiuser MIMO systems[J]. IEEE Transactions on Communications, 2013,61(4): 1436-1449.
[10]	文凯, 陈永丽, 颜飙 . 一种毫米波蜂窝网络中的D2D资源分配方案[J]. 计算机工程, 2018,44(11): 90-94.
	WEN K , CHEN Y L , YAN B . A D2D resource allocation scheme in millimeter wave cellular network[J]. Computer Engineering, 2018,44(11): 90-94.
[11]	张芳, 邹卫霞, 杜光龙 ,等. 60GHz 毫米波无线网络中基于归一化分阶效用函数功率控制算法[J]. 电子与信息学报, 2013,35(2): 291-296.
	ZHANG F , ZOU W X , DU G L ,et al. Power control based on unitary step utility function in millimeter wave WPANs[J]. Journal of Electronics ＆ Information Technology, 2013,35(2): 291-296.
[12]	施钊, 孙长印, 江帆 . 毫米波网络中基于Q-Learning的阻塞感知功率分配[J]. 计算机工程, 2020,46(12): 185-192.
	SHI Z , SUN C Y , JIANG F . Block-aware power allocation based on Q-learning in millimeter-wave network[J]. Computer Engineering, 2020,46(12): 185-192.
[13]	ROH W , SEOL J Y , PARK J ,et al. Millimeter-wave beamforming as an enabling technology for 5G cellular communications:theoretical feasibility and prototype results[J]. IEEE Communications Magazine, 2014,52(2): 106-113.
[14]	PARKVALL S , DAHLMAN E , FURUSKAR A ,et al. NR:the new 5G radio access technology[J]. IEEE Communications Standards Magazine, 2017,1(4): 24-30.
[15]	GONZALEZ-PRELCIC N , ALI A , VA V ,et al. Millimeter-wave communication with out-of-band information[J]. IEEE Communications Magazine, 2017,55(12): 140-146.
[16]	修越, 张忠培, 赵柏睿 ,等. 基于K学习的Sub-6 GHz辅助毫米波信道信息获取[J]. 电子科技大学学报, 2020,49(3): 453-457,466.
	XIU Y , ZHANG Z P , ZHAO B R ,et al. Acquisition of channel state information based on K-learning sub-6GHz assist mmWave[J]. Journal of University of Electronic Science and Technology of China, 2020,49(3): 453-457,466.
[17]	HASHEMI M , KOKSAL C E , SHROFF N B . Out-of-band millimeter wave beamforming and communications to achieve low latency and high energy efficiency in 5G systems[J]. IEEE Transactions on Communications, 2018,66(2): 875-888.
[18]	ALRABEIAH M , ALKHATEEB A . Deep learning for mmWave beam and blockage prediction using sub-6 GHz channels[J]. IEEE Transactions on Communications, 2020,68(9): 5504-5518.
[19]	LI X F , ALKHATEEB A . Deep learning for direct hybrid precoding in millimeter wave massive MIMO systems[C]// Proceedings of 2019 53rd Asilomar Conference on Signals,Systems,and Computers. Piscataway:IEEE Press, 2019: 800-805.
[20]	LUO Z Q , ZHANG S Z . Dynamic spectrum management:complexity and duality[J]. IEEE Journal of Selected Topics in Signal Processing, 2008,2(1): 57-73.
[21]	ALRABEIAH M , ALKHATEEB A . Deep learning for TDD and FDD massive MIMO:mapping channels in space and frequency[C]// Proceedings of 2019 53rd Asilomar Conference on Signals,Systems,and Computers. Piscataway:IEEE Press, 2019: 1465-1470.
[22]	VIEIRA J , LEITINGER E , SARAJLIC M ,et al. Deep convolutional neural networks for massive MIMO fingerprint-based positioning[C]// Proceedings of 2017 IEEE 28th Annual International Symposium on Personal,Indoor,and Mobile Radio Communications (PIMRC). Piscataway:IEEE Press, 2017: 1-6.
[23]	HORNIK K , STINCHCOMBE M , WHITE H . Multilayer feedforward networks are universal approximators[J]. Neural Networks, 1989,2(5): 359-366.
[24]	ALKHATEEB A . DeepMIMO:a generic deep learning dataset for millimeter wave and massive MIMO applications[J]. arXiv Preprint,arXiv:1902.06435, 2019.
[25]	ZHANG T , ZHU K , NIYATO D . A generative adversarial learning-based approach for cell outage detection in self-organizing cellular networks[J]. IEEE Wireless Communications Letters, 2020,9(2): 171-174.
[26]	徐勇军, 胡圆, 李国权 ,等. 异构携能通信网络顽健资源分配算法[J]. 通信学报, 2019,40(7): 186-196.
	XU Y J , HU Y , LI G Q ,et al. Robust resource allocation algorithm for heterogeneous wireless network with SWIPT[J]. Journal on Communications, 2019,40(7): 186-196.

参数	值
隐藏层数/个	3
隐藏层节点数/个	200,80,80
优化器	RMSprop
学习率	2×10^-3
训练期数	500
批大小	300
数据集分割	0.8/0.2

参数	场景1	场景2	场景3
激活基站序号	[1,2,3,4]	[5,6,7,8]	[1,2,3,4]
带宽（Sub-6/mmWave）/GHz^[18]	0.5/0.02	0.5/0.02	0.5/0.02
噪声功率/(dBm·MHz^-1)^[18]	-134	-134	-134
天线数目（Sub-6/mmWave）/个	4/16	4/16	4/16
用户候选位置数/个	约1.1×10⁵	9.7×10⁴	1×10⁴
训练集样本数/个	5×10⁵	5×10⁵	5×10⁵
测试集样本数/个	5×10⁴	5×10⁴	5×10⁴

算法	场景1平均和速率/(bit·s^-1)	场景2平均和速率/(bit·s^-1)
Sub6-DNN	8.700	8.406
mmW-DNN	8.702	8.411
WMMSE	8.725	8.484

算法	平均和速率/(bit·s^-1)	耗时/s
Sub6-DNN	11.491	0.210
WMMSE	11.676	632.232
比值	98.417%	0.033%

算法	平均和速率/(bit·s^-1)	耗时/s
Sub6-DNN	11.536	0.380
WMMSE	11.814	1 381.035
比值	97.652%	0.028%

基于DNN的Sub-6 GHz辅助毫米波网络功率分配算法

DNN-based Sub-6 GHz assisted millimeter wave network power allocation algorithm

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 12

参考文献 26

相关文章 15

Metrics

推荐阅读 0

[1]	陈东昱, 陈华, 范丽敏, 付一方, 王舰. 基于深度学习的随机性检验策略研究[J]. 通信学报, 2023, 44(6): 23-33.
[2]	李荣鹏, 汪丙炎, 张宏纲, 赵志峰. 知识增强的语义通信接收端设计[J]. 通信学报, 2023, 44(6): 70-76.
[3]	马帅, 裴科, 祁华艳, 李航, 曹雯, 王洪梅, 熊海良, 李世银. 基于生成模型的地磁室内高精度定位算法研究[J]. 通信学报, 2023, 44(6): 211-222.
[4]	陈晋音, 熊海洋, 马浩男, 郑雅羽. 基于对比学习的图神经网络后门攻击防御方法[J]. 通信学报, 2023, 44(4): 154-166.
[5]	李建锋, 刘哲宇, 荣洋, 李展, 廖柏林, 屈林曦, 刘志杰, 林琨煌. 用于线性噪声时变凸二次规划的归零神经网络[J]. 通信学报, 2023, 44(4): 226-233.
[6]	林云, 徐怀韬, 王森, 张思成, 庄龙. 基于特征融合的通信语音干扰效果客观评估[J]. 通信学报, 2023, 44(3): 105-116.
[7]	杨宏宇, 杨海云, 张良, 成翔. 基于特征依赖图的源代码漏洞检测方法[J]. 通信学报, 2023, 44(1): 103-117.
[8]	宋长庆, 赵宏志, 秦俪之, 邵士海. 频相失配下跳频保密通信性能分析与功率优化[J]. 通信学报, 2022, 43(9): 42-56.
[9]	马帅, 李兵, 盛海鸿, 谷荣妍, 周辉, 王洪梅, 王悦, 李世银. 基于深度强化学习的可见光定位通信一体化功率分配研究[J]. 通信学报, 2022, 43(8): 121-130.
[10]	何世文, 袁军, 安振宇, 张敏, 黄永明, 张尧学. 基于图神经网络的联合用户调度与波束成形优化算法[J]. 通信学报, 2022, 43(7): 73-84.
[11]	杨洁, 董标, 付雪, 王禹, 桂冠. 基于轻量化分布式学习的自动调制分类方法[J]. 通信学报, 2022, 43(7): 134-142.
[12]	冷涛, 蔡利君, 于爱民, 朱子元, 马建刚, 李超飞, 牛瑞丞, 孟丹. 基于系统溯源图的威胁发现与取证分析综述[J]. 通信学报, 2022, 43(7): 172-188.
[13]	孙钢灿, 吴新李, 郝万明, 朱政宇. 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[J]. 通信学报, 2022, 43(6): 179-188.
[14]	李昂, 陈建新, 魏昕, 周亮. 面向6G的跨模态信号重建技术[J]. 通信学报, 2022, 43(6): 28-40.
[15]	杨秀璋, 彭国军, 李子川, 吕杨琦, 刘思德, 李晨光. 基于Bert和BiLSTM-CRF的APT攻击实体识别及对齐研究[J]. 通信学报, 2022, 43(6): 58-70.