基于二维异步同相正交直方图的调制方式识别

doi:10.19665/j.issn1001-2400.20240312

摘要/Abstract

摘要：

自动调制识别技术能够准确识别信号的调制类型,是信号处理领域中的一项关键技术。传统的识别方法在低信噪比下存在着识别准确率低的问题,并且当信号频率不稳定,或存在异步采样时,常规的识别方法性能将会出现恶化甚至失效。基于此,在信噪比和延迟变化的信道条件下,针对低速异步采样信号,研究基于深度学习的调制方式识别技术。首先对低速异步采样信号进行建模,利用其同相正交分量生成二维异步同相正交直方图,然后通过径向基函数神经网络提取该二维图像的特征参数,完成对输入信号的调制方式识别,最后经过大量的计算机仿真验证了该方法对7种调制方式的识别准确率。实验结果表明,在受到加性高斯白噪声的信道模型下,在低速异步采样的输入信号信噪比为6 dB时,可达到约95%以上的平均识别准确率,并通过对比实验验证了所提方案的有效性和鲁棒性。

关键词: 调制方式识别, 二维异步同相正交直方图, 深度学习, 径向基神经网络

Abstract:

Automatic modulation recognition technology accurately identifies the modulation type of signals,making it a key technology in the field of signal processing.Traditional recognition methods suffer from low accuracy at low signal-to-noise ratios,and performance degradation or failure when dealing with signal frequency instability or asynchronous sampling.In this paper,we investigate modulation recognition technology based on deep learning for low-speed asynchronous sampled signals under channel conditions with varying signal-to-noise ratios and delays.We start by modeling low-speed asynchronous sampled signals and generating a two-dimensional asynchronous in-phase quadrature histogram using their in-phase and quadrature components.Subsequently,we employ a Radial Basis Function Neural Network to extract feature parameters from this two-dimensional image,thus achieving modulation type recognition for the input signal.Extensive computer simulations validate the proposed method’s accuracy in recognizing seven modulation types under the influence of additive white Gaussian noise.Experimental results demonstrate that,in the presence of additive white Gaussian noise in the channel model and with an input signal-to-noise ratio of 6 dB,the average recognition accuracy can exceed 95%.Comparative experiments further verify the effectiveness and robustness of the proposed approach.

Key words: modulation recognition, two-dimensional asynchronous in-phase quadrature histogram, deep learning, radial basis function neural network

中图分类号:

TN911.6

万鹏武, 惠茜, 陈东瑞, 吴波. 基于二维异步同相正交直方图的调制方式识别[J]. 西安电子科技大学学报, 2024, 51(4): 78-90.

WAN Pengwu, HUI Xi, CHEN Dongrui, WU Bo. Modulation recognition based on the two-dimensional asynchronous in-phase quadrature histogram[J]. Journal of Xidian University, 2024, 51(4): 78-90.

图/表 12

图1

图2

图3

图4

表1

表2

图5

图6

图7

图8

图9

表3

参考文献 31

[1]	SHI Q, KARASAWA Y. Automatic Modulation Identification Based on the Probability Density Function of Signal Phase[J]. IEEE Transactions on Communications, 2012, 60(4):1033-1044.
[2]	JDID B, HASSAN K, DAYOUB I, et al. Machine Learning Based Automatic Modulation Recognition for Wireless Com-Munications:A Comprehensive Survey[J]. IEEE Access, 2021,9:57851-57873.
[3]	WEN W, MENDEL J M. Maximum-Likelihood Classification for Digital Amplitude-Phase Modulations[J]. IEEE Transactions on Communications, 2000, 48(2):189-193.
[4]	ZHANG R, HE C, JING L, et al. A Modulation Recognition System for Underwater Acoustic Communication Signals Based on Higher-Order Cumulants and Deep Learning[J]. Journal of Marine Science and Engineering, 2023, 11(8):1632.
[5]	SHAH M H, DANG X. Classification of Spectrally Efficient Constant Envelope Modulations Based on Radial Basis Function Network and Deep Learning[J]. IEEE Communications Letters, 2019, 23(9):1529-1533.
[6]	SHI W, LIU D, CHENG X, et al. Particle Swarm Optimization-Based Deep Neural Network for Digital Modulation Recognition[J]. IEEE Access, 2019,7:104591-104600.
[7]	刘高高, 黄东杰, 席昕, 等. 一种特征融合的工作模式识别方法[J]. 西安电子科技大学学报, 2023, 50(6):13-20.
	LIU Gaogao, HUANG Dongjie, XI Xin, et al. Work Pattern Recognition Method Based on Feature Fusion[J]. Journal of Xidian University, 2023, 50(6):13-20.
[8]	QU W, YAO G, MENG L. Research on Radar PRI Modulation Pattern Recognition Based on Recurrent Neural Network[C]//2023 4th International Conference on Computer Vision,Image and Deep Learning (CVIDL).Piscataway:IEEE, 2023: 250-254.
[9]	ZHANG H, LI T, LI Y, et al. Virtual Electromagnetic Environment Modeling Based Data Augmentation for Drone Signal Identification[J]. Journal of Information and Intelligence, 2023, 1(4):308-320.
[10]	杨静雅, 齐彦丽, 周一青, 等. CNN-Transformer轻量级智能调制识别算法[J]. 西安电子科技大学学报, 2023, 50(3):40-49.
	YANG Jingya, QI Yanli, ZHOU Yiqing, et al. Algorithm for Recognition of Lightweight Intelligent Modulation Based on the CNN-Transformer Networks[J]. Journal of Xidian University, 2023, 50(3):40-49.
[11]	SALEHIN I, ISLAM M S, SAHA P, et al. AutoML:A Systematic Review on Automated Machine Learning with Neural Architecture Search[J]. Journal of Information and Intelligence, 2023, 2(1):52-81.
[12]	TUNZE G B, HUYNH-THE T, LEE J M, et al. Sparsely Connected CNN for Efficient Automatic Modulation Recognition[J]. IEEE Transactions on Vehicular Technology, 2020, 69(12):15557-15568.
[13]	LIU M, YANG K, ZHAO N, et al. Intelligent Signal Classification in Industrial Distributed Wireless Sensor Networks Based Industrial Internet of Things[J]. IEEE Transactions on Industrial Informatics, 2020, 17(7):4946-4956.
[14]	杜明洋, 杜蒙, 潘继飞, 等. 基于生成对抗网络的雷达脉内信号去噪与识别[J]. 西安电子科技大学学报, 2023, 50(6):133-147.
	DU Mingyang, DU Meng, PAN Jifei, et al. Generative Adversarial Model for Radar Intra-Pulse Signal Denoising and Recognition[J]. Journal of Xidian University, 2023, 50(6):133-147.
[15]	KARRA K, KUZDEBA S, PETERSEN J. Modulation Recognition Using Hierarchical Deep Neural Networks[C]//2017 IEEE International Symposium on Dynamic Spectrum Access Networks(DySPAN).Piscataway:IEEE, 2017: 1-3.
[16]	LEE J H, KIM K Y, SHIN Y. Feature Image-Based Automatic Modulation Classification Method Using CNN Algorithm[C]//2019 International Conference on Artificial Intelligence in Information and Communication(ICAIIC).Piscataway:IEEE, 2019: 1-4.
[17]	HUANG X, LI X. Modulation Identification Method Based on Time-Frequency Analysis and Support Vector Machine[C]//2023 IEEE 2nd International Conference on Electrical Engineering,Big Data and Algorithms(EEBDA).Piscataway:IEEE, 2023: 551-554.
[18]	WEI S, QU Q, SU H, et al. Intra-Pulse Modulation Radar Signal Recognition Based on Squeeze-and-Excitation Networks[J]. Signal,Image and Video Processing, 2020,14:1133-1141.
[19]	O’SHEA T J, CORGAN J, CLANCY T C. Convolutional Radio Modulation Recognition Networks[C]//Engineering Applications of Neural Networks:17th International Conference,EANN 2016.Heidelberg:Springer, 2016:213-226.
[20]	HONG S, ZHANG Y, WANG Y, et al. Deep Learning-Based Signal Modulation Identification in OFDM Systems[J]. IEEE Access, 2019,7:114631-114638.
[21]	KHAN F N, ZHONG K, AL-ARASHI W H, et al. Modulation Format Identification in Coherent Receivers Using Deep Machine Learning[J]. IEEE Photonics Technology Letters, 2016, 28(17):1886-1889.
[22]	KHAN F N, ZHONG K, ZHOU X, et al. Joint OSNR Monitoring and Modulation Format Identification in Digital Coherent Receivers Using Deep Neural Networks[J]. Optics express, 2017, 25(15):17767-17776. doi: 10.1364/OE.25.017767 pmid: 28789268
[23]	YANG F, BAI C, GAO H, et al. Joint Modulation Format Identification and Mode Coupling Estimation Scheme Based on ADTP and MT-CNN for Mode Division Multiplexed Systems[C]//2022 Asia Communications and Photonics Conference (ACP).Piscataway:IEEE, 2022: 737-741.
[24]	SAIF W S, RAGHEB A M, NEBENDAHL B, et al. Performance Investigation of Modulation Format Identification in Super-Channel Optical Networks[J]. IEEE Photonics Journal, 2022, 14(2):1-10.
[25]	LEUNG T C, LEE C N. Flow-Based DDoS Detection Using Deep Neural Network with Radial Basis Function Neural Network[C]//2022 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC).Piscataway:IEEE, 2022: 1774-1779.
[26]	XIA L, HU P, MA K, et al. Research on Measurement Modeling of Spherical Joint Rotation Angle Based on RBF-ELM Network[J]. IEEE Sensors Journal, 2021, 21(20):23118-23124.
[27]	PANDA S, PANDA G. On the Development and Performance Evaluation of Improved Radial Basis Function Neural Networks[J]. IEEE Transactions on Systems,Man,and Cybernetics:Systems, 2021, 52(6):3873-3884.
[28]	CHEN B H, HUANG S C, LI C Y, et al. Haze Removal Using Radial Basis Function Networks for Visibility Restoration Applications[J]. IEEE Transactions on Neural Networks and Learning Systems, 2017, 29(8):3828-3838.
[29]	SLEDGE I J, PRÍNCIPE J C. An Exact Reformulation of Feature-Vector-Based Radial-Basis-Function Networks for Graph-Based Observations[J]. IEEE Transactions on Neural Networks and Learning Systems, 2019, 31(11):4990-4998.
[30]	ELANSARI T, OUANAN M, BOURRAY H. Mixed Radial Basis Function Neural Network Training Using Genetic Algorithm[J]. Neural Processing Letters, 2023, 55(8):10569-10587.
[31]	DEVI M G, AKILA I S. Deep Learning based Compressive Sensing-Radial Basis Functional Neural Network for Image Fusion[C]//2023 4th International Conference on Signal Processing and Communication (ICSPC).Piscataway:IEEE, 2023: 432-435.

参数	值	参数	值
符号速率/(symbols·s^-1)	4.1×10⁸	符号周期/s	2.44×10^-9
异步采样速率/(Mb·s^-1)	30	载波频率/Hz	4.1×10⁹
滤波器	滚降因子R=0.75	区间个数	40
信噪比/dB	2~16(步长为1)	信道	高斯白噪声信道
神经网络	径向基函数神经网络	总样本数	52 500
调制类型		BPSK、QPSK、8PSK、4QAM、16QAM、32QAM、64QAM

算法	训练时间
RBFNN	209.16
ELM	238.78
随机森林	309.94
SVM	336.23
CNN	655.17