基于残差密集网络的频谱感知方法

doi:10.11959/j.issn.1000-436x.2021220

通信学报 ›› 2021, Vol. 42 ›› Issue (12): 182-191.doi: 10.11959/j.issn.1000-436x.2021220

基于残差密集网络的频谱感知方法

盖建新, 薛宪峰, 南瑞祥, 吴静谊

哈尔滨理工大学测控技术与仪器黑龙江省高校重点实验室，黑龙江哈尔滨 150080

修回日期:2021-11-04 出版日期:2021-12-01 发布日期:2021-12-01
作者简介:盖建新（1980- ），男，辽宁朝阳人，博士，哈尔滨理工大学副教授，主要研究方向为频谱感知、机器学习、亚奈奎斯特采样理论、压缩感知等
薛宪峰（1996- ），男，山东菏泽人，哈尔滨理工大学硕士生，主要研究方向为深度学习
南瑞祥（1996- ），男，山东济宁人，哈尔滨理工大学硕士生，主要研究方向为通信信号处理
吴静谊（1996- ），女，黑龙江绥化人，哈尔滨理工大学硕士生，主要研究方向为压缩感知
基金资助:
国家自然科学基金资助项目(61501150);黑龙江省自然科学基金资助项目(QC2014C074);黑龙江省省属本科高校基本科研业务费基金资助项目(2018-KYYWF-1656)

Spectrum sensing method based on residual dense network

Jianxin GAI, Xianfeng XUE, Ruixiang NAN, Jingyi WU

The Higher Educational Key Laboratory for Measuring ＆Control Technology and Instrumentations of Heilongjiang Province， Harbin University of Science and Technology， Harbin 150080， China

Revised:2021-11-04 Online:2021-12-01 Published:2021-12-01
Supported by:
The National Natural Science Foundation of China(61501150);The Natural Science Foundation of Heilong-jiang Province(QC2014C074);The Fundamental Research Funds for the Universities in Heilongjiang Province(2018-KYYWF-1656)

摘要/Abstract

摘要：

针对传统卷积神经网络（CNN）频谱感知方法没有充分利用特征图信息并且提取特征图的能力受限于浅层的网络结构等问题，通过在传统 CNN 频谱感知方法中添加密集连接，实现特征图信息重利用，同时在密集单元的两端加入捷径连接，实现更深层的网络训练，进而提出一种基于残差密集网络（ResDenNet）的频谱感知方法。该方法将频谱感知问题映射为图像二分类问题，首先对接收信号分割成矩阵并归一化灰度处理，得到的灰度图像作为网络的输入，然后通过密集学习和残差学习训练网络，最后将在线数据输入ResDenNet中，完成基于图像分类的频谱感知。数值实验表明，所提方法优于传统频谱感知方法，在信噪比低至-19 dB时，所提方法检测概率仍高达0.96，虚警概率低至0.1，同时具有更好的泛化能力。

关键词: 频谱感知, 残差密集网络, 密集连接, 捷径连接

Abstract:

Aiming at the problem that the traditional spectrum sensing method based on convolutional neural network（CNN） did not make full use of image feature and the ability of extracting the image feature was limited by the shallow network structure， a spectrum sensing method based on the residual dense network （ResDenNet） was proposed.By adding dense connections in the traditional neural network， the information reuse of the image feature was achieved.Meanwhile， shortcut connections were employed at both ends of the dense unit to implement deeper network training.The spectrum sensing problem was transformed into the image binary classification problem.Firstly， the received signals were integrated into a matrix， which was normalized and transformed by gray level.The obtained gray level images were used as the input of the network.Then， the network was trained through dense learning and residual learning.Finally， the online data was input into the ResDenNet and spectrum sensing was implemented based on image classification.The numerical experiments show that the proposed method is superior to the traditional ones in terms of performance.When the SNR is as low as -19 dB， the detection probability of the proposed method is still high up to 0.96 with a low false alarm probability of 0.1， while a better generalization ability is displayed.

Key words: spectrum sensing, ResDenNet, dense connection, shortcut connection

中图分类号:

TN911

盖建新, 薛宪峰, 南瑞祥, 吴静谊. 基于残差密集网络的频谱感知方法[J]. 通信学报, 2021, 42(12): 182-191.

Jianxin GAI, Xianfeng XUE, Ruixiang NAN, Jingyi WU. Spectrum sensing method based on residual dense network[J]. Journal on Communications, 2021, 42(12): 182-191.

图/表 14

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

表1

图11

图12

图13

参考文献 25

[1]	DIGHAM F F , ALOUINI M S , SIMON M K . On the energy detection of unknown signals over fading channels[J]. IEEE Transactions on Communications, 2007,55(1): 21-24.
[2]	WU J Y , WANG C H , WANG T Y . Performance analysis of energy detection based spectrum sensing with unknown primary signal arrival time[J]. IEEE Transactions on Communications, 2011,59(7): 199-206.
[3]	OH H , NAM H . Energy detection scheme in the presence of burst signals[J]. IEEE Signal Processing Letters, 2019,26(4): 582-586.
[4]	YANG M C , LI Y , LIU X F ,et al. Cyclostationary feature detection based spectrum sensing algorithm under complicated electromagnetic environment in cognitive radio networks[J]. China Communications, 2015,12(9): 35-44.
[5]	SHEN J C , ALSUSA E . Joint cycle frequencies and lags utilization in cyclostationary feature spectrum sensing[J]. IEEE Transactions on Signal Processing, 2013,61(21): 5337-5346.
[6]	ZHANG X Z , GAO F F , CHAI R ,et al. Matched filter based spectrum sensing when primary user has multiple power levels[J]. China Communications, 2015,12(2): 21-31.
[7]	LIU C , WANG J , LIU X M ,et al. Maximum eigenvalue-based goodness-of-fit detection for spectrum sensing in cognitive radio[J]. IEEE Transactions on Vehicular Technology, 2019,68(8): 7747-7760.
[8]	AWE O P , DELIGIANNIS A , LAMBOTHARAN S . Spatio-temporal spectrum sensing in cognitive radio networks using beamformer-aided SVM algorithms[J]. IEEE Access, 2018,6: 25377-25388.
[9]	BAO J R , NIE J Y , LIU C ,et al. Improved blind spectrum sensing by covariance matrix Cholesky decomposition and RBF-SVM decision classification at low SNRs[J]. IEEE Access, 2019,7: 97117-97129.
[10]	陈思吉, 王欣, 申滨 . 一种基于支持向量机的认知无线电频谱感知方案[J]. 重庆邮电大学学报(自然科学版), 2019,31(3): 313-322.
	CHEN S J , WANG X , SHEN B . A support vector machine based spectrum sensing for cognitive radios[J]. Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition), 2019,31(3): 313-322.
[11]	TANG Y J , ZHANG Q Y , LIN W . Artificial neural network based spectrum sensing method for cognitive radio[C]// Proceedings of 2010 6th International Conference on Wireless Communications Networking and Mobile Computing (WiCOM). Piscataway:IEEE Press, 2010: 1-4.
[12]	VYAS M R , PATEL D K , LOPEZ-BENITEZ M , . Artificial neural network based hybrid spectrum sensing scheme for cognitive radio[C]// Proceedings of 2017 IEEE 28th Annual International Symposium on Personal,Indoor,and Mobile Radio Communications (PIMRC). Piscataway:IEEE Press, 2017: 1-7.
[13]	高红民, 曹雪莹, 杨耀 ,等. 基于CNN的双边融合网络在高光谱图像分类中的应用[J]. 通信学报, 2020,41(11): 132-140.
	GAO H M , CAO X Y , YANG Y ,et al. Application of bilateral fusion model based on CNN in hyperspectral image classification[J]. Journal on Communications, 2020,41(11): 132-140.
[14]	YU C Y , HAN R , SONG M P ,et al. A simplified 2D-3D CNN architecture for hyperspectral image classification based on spatial-spectral fusion[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2020,13: 2485-2501.
[15]	MAFFEI A , HAUT J M , PAOLETTI M E ,et al. A single model CNN for hyperspectral image denoising[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020,58(4): 2516-2529.
[16]	CHEN Y S , ZHU K Q , ZHU L ,et al. Automatic design of convolutional neural network for hyperspectral image classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019,57(9): 7048-7066.
[17]	LIU C , WANG J , LIU X M ,et al. Deep CM-CNN for spectrum sensing in cognitive radio[J]. IEEE Journal on Selected Areas in Communications, 2019,37(10): 2306-2321.
[18]	LEE W , KIM M , CHO D H . Deep cooperative sensing:cooperative spectrum sensing based on convolutional neural networks[J]. IEEE Transactions on Vehicular Technology, 2019,68(3): 3005-3009.
[19]	张孟伯, 王伦文, 冯彦卿 . 基于卷积神经网络的OFDM频谱感知方法[J]. 系统工程与电子技术, 2019,41(1): 178-186.
	ZHANG M B , WANG L W , FENG Y Q . OFDM spectrum sensing method based on convolutional neural networks[J]. Systems Engineering and Electronics, 2019,41(1): 178-186.
[20]	MARSHALL J A . Neural networks for pattern recognition[J]. Neural Networks, 1995,8(3): 493-494.
[21]	JAMES D A , VENABLES W N , RIPLEY B D . Modern applied statistics with S-PLUS[J]. Technometrics, 1996,38(1): 77.
[22]	PHATAK D S , KOREN I . Connectivity and performance tradeoffs in the cascade correlation learning architecture[J]. IEEE Transactions on Neural Networks, 1994,5(6): 930-935.
[23]	WILAMOWSKI B M , YU H . Neural network learning without backpropagation[J]. IEEE Transactions on Neural Networks, 2010,21(11): 1793-1803.
[24]	陈丽, 陈静, 高新涛 ,等. 基于支持向量机与反 K 近邻的分类算法研究[J]. 计算机工程与应用, 2010,46(24): 135-137,188.
	CHEN L , CHEN J , GAO X T ,et al. Classification algorithm research based on support vector machine and reverse K-nearest neighbor[J]. Computer Engineering and Applications, 2010,46(24): 135-137,188.
[25]	孙月驰, 李冠 . 基于卷积神经网络嵌套模型的人群异常行为检测[J]. 计算机应用与软件, 2019,36(3): 196-201,276.
	SUN Y C , LI G . Abnormal behavior detection of crowds based on nested model of convolutional neural network[J]. Computer Applications and Software, 2019,36(3): 196-201,276.

算法	P_d	P_fa	训练时间/s	感知时间/s
ResDenNet	0.94	0.06	32.42	4.06
CNN-5L	0.83	0.19	25.66	2.80
CNN-20L	0.55	0.14	36.11	4.62
SVM	0.42	0.36	15.24	10.72

基于残差密集网络的频谱感知方法

Spectrum sensing method based on residual dense network

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 25

相关文章 15

Metrics

推荐阅读 0

[1]	钱志鸿, 肖琳, 王雪. 面向未来移动网络密集连接的关键技术综述[J]. 通信学报, 2021, 42(4): 22-43.
[2]	孙志国, 任欣悦, 陈增茂, 刁鸣. 基于证据间相似性的协作频谱感知方法与性能分析[J]. 通信学报, 2020, 41(12): 139-147.
[3]	赵文静,李贺,金明录. 基于特征值的频谱感知融合算法[J]. 通信学报, 2019, 40(11): 57-64.
[4]	付元华,贺知明. 协作频谱感知中基于距离准则的量化器设计[J]. 通信学报, 2018, 39(9): 49-56.
[5]	刘玉磊,梁俊,肖楠,袁小刚,张振浩. 基于信道历史状态感知信息的频谱感知方法[J]. 通信学报, 2017, 38(8): 118-130.
[6]	余盼,李斌,赵成林. 基于能量检测的异步感知算法[J]. 通信学报, 2017, 38(3): 165-173.
[7]	冯景瑜,杜续,王宏刚,黄文华. 认知ad hoc网络中的多信道路由安全威胁及其对策研究[J]. 通信学报, 2016, 37(Z1): 92-97.
[8]	张洋,彭华,巩克现. 基于多尺度功率谱子带梯度的宽带频谱感知算法与性能分析[J]. 通信学报, 2016, 37(2): 191-198.
[9]	刘畅,张锐,李思莹,王洁,金明录. 基于空间谱的多天线盲频谱感知算法[J]. 通信学报, 2015, 36(4): 115-124.
[10]	弥寅,卢光跃. 基于特征值极限分布的合作频谱感知算法[J]. 通信学报, 2015, 36(1): 84-89.
[11]	左旭舟,夏玮玮,沈连丰. LTE系统中femtocell的动态频谱接入方法[J]. 通信学报, 2015, 36(1): 111-120.
[12]	孙梦巍,赵龙,许巧春,李斌,赵成林. 动态时变衰落信道下的频谱感知算法[J]. 通信学报, 2014, 35(7): 63-69.
[13]	孙梦巍，赵龙，许巧春，李斌，赵成林. 一种动态时变衰落信道下的频谱感知算法[J]. 通信学报, 2014, 35(7): 8-69.
[14]	申滨，王舒，黄琼，陈前斌. 基于Gerschgorin圆盘理论的认知无线电宽带频谱感知[J]. 通信学报, 2014, 35(4): 1-10.
[15]	申滨,王舒,黄琼,陈前斌. 基于Gerschgorin圆盘理论的认知无线电宽带频谱感知[J]. 通信学报, 2014, 35(4): 1-10.