基于深层神经网络的信道编码类型盲识别

doi:10.12305/j.issn.1001-506X.2024.05.35

系统工程与电子技术 ›› 2024, Vol. 46 ›› Issue (5): 1820-1829.doi: 10.12305/j.issn.1001-506X.2024.05.35

• 通信与网络 • 上一篇

基于深层神经网络的信道编码类型盲识别

杨宗方¹^,²^,*, 张天骐¹^,², 马焜然¹^,², 邹涵¹^,²

1. 重庆邮电大学通信与信息工程学院, 重庆 400065
2. 重庆邮电大学信号与信息处理重庆市重点实验室, 重庆 400065

收稿日期:2023-02-20 出版日期:2024-04-30 发布日期:2024-04-30
通讯作者: 杨宗方
作者简介:杨宗方(1998—), 男, 硕士研究生, 主要研究方向为信道编码盲识别技术研究
张天骐(1971—), 男, 教授, 博士, 主要研究方向为通信信号的调制解调与盲处理、图像与语音信号处理、神经网络实现、现场可编程逻辑门阵列实现和超大规模集成电路实现
马焜然(1999—), 男, 硕士研究生, 主要研究方向为数字水印、信息隐藏
邹涵(1998—), 男, 硕士研究生, 主要研究方向为通信信号盲处理
基金资助:
国家自然科学基金(61671095);国家自然科学基金(61702065);国家自然科学基金(61701067);国家自然科学基金(61771085);信号与信息处理重庆市市级重点实验室建设项目(CSTC2009CA2003);重庆市自然基金(cstc2021jcyj-msxmX0836);重庆市教育委员会科研项目(KJ1600427);重庆市教育委员会科研项目(KJ1600429)

Blind identification of channel coding types based on deep neural networks

Zongfang YANG¹^,²^,*, Tianqi ZHANG¹^,², Kunran MA¹^,², Han ZOU¹^,²

1. School of Communication and Information Engineering, Chongqing University of Posts and Telecommunications, Chongqing 400065, China
2. Chongqing Key Laboratory of Signal and Information Processing, Chongqing University of Posts and Telecommunications, Chongqing 400065, China

Received:2023-02-20 Online:2024-04-30 Published:2024-04-30
Contact: Zongfang YANG

摘要/Abstract

摘要：

为了解决当前识别算法只能识别一种或者两种码字类型以及人工提取特征复杂的问题, 提出了两种基于深层神经网络模型的信道编码类型识别器, 即卷积神经网络(convolutional neural network, CNN)识别器和递归CNN (recursive CNN, RCNN)识别器, 用于识别接收数据中不同类型的信道码字。将待识别的软解调序列作为自然语言处理中文本分类问题的句子向量进行处理, 输入到预先训练好的深层神经网络识别器中进行识别, 并分析了字长度对识别准确率的影响, 得出了最合适的字长度。实验结果表明, 两种识别器都能够有效识别接收数据中多种类型的信道编码, 且在信噪比为3 dB时CNN识别器的识别准确率能够达到99%以上, 而RCNN识别器在1 dB时就能够达到99%以上的识别准确率。

关键词: 深层神经网络, 信道编码识别器, 盲识别, 字长度

Abstract:

In order to solve the problem that the current recognition algorithm can only recognize one or two code types and the complexity of manually extracting features, two channel coding type recognizers based on the deep neural network model are proposed, namely, convolutional neural network (CNN) recognizer and recursive CNN (RCNN) recognizer, used to identify different types of channel codewords in received data. The soft demodulation sequence to be recognized is treated as the sentence vector of text classification in natural language processing, input into the pre trained deep neural network recognizer for recognition, and analyze the influence of word length on recognition accuracy, and obtain the most appropriate word length. The experimental results show that both types of recognizers can effectively recognize various types of channel codes in the received data, and the recognition accuracy of the CNN recognizer can reach over 99% when the signal-to-noise ratio is 3 dB, while the RCNN recognizer can achieve over 99% recognition accuracy at 1 dB.

Key words: deep neural networks, channel code recognizer, blind identification, word length

中图分类号:

TN911.7

杨宗方, 张天骐, 马焜然, 邹涵. 基于深层神经网络的信道编码类型盲识别[J]. 系统工程与电子技术, 2024, 46(5): 1820-1829.

Zongfang YANG, Tianqi ZHANG, Kunran MA, Han ZOU. Blind identification of channel coding types based on deep neural networks[J]. Systems Engineering and Electronics, 2024, 46(5): 1820-1829.

图/表 16

图1

图2

图3

图4

表1

图5

表2

表3

表4

图6

图7

表5

图8

图9

图10

表6

参考文献 20

1	闫华. 信道编码盲识别技术研究[D]. 长沙: 国防科技大学, 2018.
	YAN H. Research on blind identification technology of channel coding[D]. Changsha: National University of Defense Technology, 2018.
2	解辉,黄知涛,王丰华.信道编码盲识别技术进展[J].电子学报,2013,41(6):1166-1176.
	XIEH,HUANGZ T,WANGF H.Research progress of channel coding blind identification technology[J].Journal of Electronics,2013,41(6):1166-1176.
3	YANG S, TANG Y. Text classification based on convolutional neural network and attention model[C]//Proc. of the International Conference on Artificial Intelligence and Big Data, 2020: 67-73.
4	XIAT,WUH C.Novel blind identification of LDPC codes using average LLR of syndrome a posteriori probability[J].IEEE Trans.on Signal Processing,2014,62(3):632-640. doi: 10.1109/TSP.2013.2293975
5	于沛东,彭华,巩克现,等.基于寻找小重量码字算法的LDPC码开集识别[J].通信学报,2017,38(6):108-117.
	YUP D,PENGH,GONGK X,et al.LDPC code open set reco-gnition based on searching for low-weight codeword algorithm[J].Journal of Communications,2017,38(6):108-117.
6	WU G, ZHANG B N, WEN X, et al. Blind recognition of BCH code based on Galois field Fourier transform[C]//Proc. of the International Conference on Wireless Communications & Signal Processing, 2015.
7	王垚,王聪,王翔,等.一种基于软判决的非删余极化码参数识别算法[J].系统工程与电子技术,2023,45(10):3291-3299.
	WANGY,WANGC,WANGX,et al.A parameter recognition algorithm for non censored polarization codes based on soft decision[J].Systems Engineering and Electronic Technology,2023,45(10):3291-3299.
8	刘鉴兴,张天骐,柏浩钧,等.基于信息矩阵估计的极化码参数盲识别算法[J].系统工程与电子技术,2022,44(2):668-676.
	LIUJ X,ZHANGT Q,BAIH J,et al.Blind identification algorithm for polarization code parameters based on information matrix estimation[J].Systems Engineering and Electronic Technology,2022,44(2):668-676.
9	SWAMINATHANR,MADHUKUMARA S.Classification of error correcting codes and estimation of interleaver parameters in a noisy transmission environment[J].IEEE Trans.on Broadcasting,2017,63(3):463-478. doi: 10.1109/TBC.2017.2704436
10	陶志勇,闫明豪,刘影.基于时序卷积网络的信道编码闭集识别[J].华中科技大学学报(自然科学版),2022,50(3):12-17.
	TAOZ Y,YANM H,LIUY.Channel coding closed set reco-gnition based on temporal convolutional network[J].Journal of Huazhong University of Science and Technology (Natural Science Edition),2022,50(3):12-17.
11	SHENB X,HUANGC,XUW J,et al.Blind channel codes recognition via deep learning[J].IEEE Journal on Selected Areas in Communications,2021,39(8):2421-2433.
12	SHEN B X, WU H Y, HUANG C. Blind recognition of channel codes via deep learning[C]//Proc. of the IEEE Global Conference on Signal and Information Processing, 2019.
13	HUANG X R, SUN S J, YANG X, et al. Recognition of channel codes based on BiLSTM-CNN[C]//Proc. of the 31st Wireless and Optical Communications Conference, 2022: 151-154.
14	MEIF,CHENH,LEIY K.Blind recognition of forward error correction codes based on recurrent neural network[J].Sensors,2021,21(11):3884.
15	QIN X F, PENG S L, YANG X, et al. Deep learning based channel code identification using TextCNN[C]//Proc. of the IEEE International Symposium on Dynamic Spectrum Access Networks, 2019.
16	LIS D,ZHOUJ,HUANGZ P,et al.Recognition of error correcting codes based on CNN with block mechanism and embedding[J].Digital Signal Processing,2021,111,102986.
17	ARIKAN E, TELATAR E. On the rate of channel polarization[C]//Proc. of the IEEE International Symposium on Information Theory, 2009: 1493-1495.
18	ARIKANE.Channel polarization: a method for constructing capacity-achieving codes for symmetric binary-input memoryless channels[J].IEEE Trans.on Information Theory,2009,55(7):3051-3073.
19	WOODARDJ P,HANZOL.Comparative study of turbo decoding techniques: an overview[J].IEEE Trans.on Vehicular Technology,2000,49(6):2208-2233.
20	LAI S W, XU L H, LIU K, et al. Recurrent convolutional neural networks for text classification[C]//Proc. of the AAAI Conference on Artificial Intelligence, 2015.

CNN识别器	输出维度
嵌入层	嵌入维度: 128 字长度: 4
卷积层	卷积核高度: 2、3、4、5、6 卷积核宽度: 128 卷积核数量: 256

RCNN识别器	参数
嵌入层	嵌入维度: 128 字长度: 8
LSTM	LSTM尺寸: 128 LSTM层数: 1
池化层	滤波器尺寸: 68 滑动步长: 68

类型	超参数
优化器	Adam
学习率	0.001
批量大小	128
CNN激活函数	ReLU
RCNN激活函数	tanh
dropout率	0.5

信道编码类型	编码参数
LDPC码	256码长, 1/2码率 256码长, 1/3码率512 码长, 1/2码率512 码长, 1/3码率
卷积码	1/2码率, 约束长度为5 1/2码率, 约束长度为7 1/3码率, 约束长度为5 1/3码率, 约束长度为7
极化码	64码长, 1/2码率 128码长, 1/2码率 256码长, 1/2码率 512码长, 1/2码率
Turbo码	1/2码率, (7, 5) 1/2码率, (31, 17) 1/3码率, (7, 5) 1/3码率, (31, 17)

字长度L	CNN		RCNN		读数据/s
字长度L	训练/ms	测试/ms	训练/ms	测试/ms	读数据/s
1	57.34	9.40	201.1	23.76	183
2	29.92	4.69	100.2	10.28	166
3	20.73	3.25	41.79	7.46	105
4	15.00	2.52	33.75	5.36	91
5	11.81	1.89	27.73	4.44	85
6	10.23	1.67	25.39	3.58	80
7	8.93	1.45	19.61	3.06	74
8	7.81	1.24	16.02	2.52	65
9	6.48	1.12	13.75	2.24	22
10	3.28	0.93	12.50	2.04	21

基于深层神经网络的信道编码类型盲识别

Blind identification of channel coding types based on deep neural networks

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 16

参考文献 20

相关文章 11

编辑推荐

Metrics

本文评价

识别器	训练/ms	测试/ms	读数据/s
本文CNN	15.00	2.52	91
本文RCNN	16.02	2.52	65
文献[12]RNN	8.73	1.42	47
文献[12]注意力	21.61	3.52	47
文献[13]	27.33	4.45	47
文献[14]	22.01	3.63	47

[1]	刘鉴兴, 张天骐, 柏浩钧, 叶绍鹏. 基于信息矩阵估计的极化码参数盲识别算法[J]. 系统工程与电子技术, 2022, 44(2): 668-676.
[2]	于柯远, 张立民, 闫文君, 金堃. 基于深度学习的多STBC盲识别算法[J]. 系统工程与电子技术, 2021, 43(4): 1110-1118.
[3]	吴昭军, 张立民, 钟兆根. 低信噪比下卷积交织器识别[J]. 系统工程与电子技术, 2021, 43(2): 546-554.
[4]	闫文君, 张聿远, 凌青, 于柯远, 谭凯文, 刘恒燕. 基于频域互相关序列与峰值检测的空频分组码盲识别算法[J]. 系统工程与电子技术, 2021, 43(12): 3709-3715.
[5]	张聿远, 闫文君, 张立民, 张媛. 基于卷积神经网络的串行空时分组码盲识别算法[J]. 系统工程与电子技术, 2021, 43(11): 3360-3370.
[6]	凌青, 闫文君, 张立民, 于柯远. AL-OFDM和SM-OFDM空频分组码信号盲识别方法[J]. 系统工程与电子技术, 2021, 43(1): 19-25.
[7]	张天骐, 范聪聪, 喻盛琪, 赵健根. 基于JADE与特征提取的正交/非正交空时分组码盲识别[J]. 系统工程与电子技术, 2020, 42(4): 933-939.
[8]	龙浪, 杨俊安, 刘辉, 梁宗伟. 基于分析矩阵零均值比的CDL卷积交织盲识别[J]. 系统工程与电子技术, 2018, 40(8): 1702-1707.
[9]	凌青, 张立民, 闫文君, 于柯远. 频率选择信道下STBC-OFDM信号盲识别[J]. 系统工程与电子技术, 2017, 39(5): 1141-1147.
[10]	张旻, 陆凯, 李歆昊, 吕全通. 归零Turbo码的盲识别方法[J]. 系统工程与电子技术, 2016, 38(6): 1423-1428.
[11]	王丰华，解辉，黄知涛，刘晓光. 基于频谱累积量的线性分组码检测识别方法[J]. 系统工程与电子技术, 2013, 35(12): 2595-2599.