基于余弦校验关系的卷积神经网络LDPC码盲识别

doi:10.12305/j.issn.1001-506X.2025.09.34

系统工程与电子技术 ›› 2025, Vol. 47 ›› Issue (9): 3117-3125.doi: 10.12305/j.issn.1001-506X.2025.09.34

• 通信与网络 • 上一篇

基于余弦校验关系的卷积神经网络LDPC码盲识别

陈文洁¹(), 张浦²^,*, 史高翔³, 刘林², 刘烜²

1. 中国电子科技集团第十研究所，四川成都 610036
2. 西南交通大学信息编码与传输省重点实验室，四川成都 611756
3. 93147部队，四川成都 610041

收稿日期:2024-08-05 出版日期:2025-09-25 发布日期:2025-09-16
通讯作者: 张浦 E-mail:chenwenjiespring@163.com
作者简介:陈文洁（1989—），女，高级工程师，硕士，主要研究方向为通信信号处理与分析、信号智能处理
史高翔（1989—），男，工程师，本科，主要研究方向为信号通信系统分析
刘　林（1974—），女，副教授，博士，主要研究方向为信号处理、通信对抗、无线定位
刘　烜（2000—），男，硕士研究生，主要研究方向为信道译码、信号处理、通信对抗
基金资助:
中电天奥有限公司创新理论技术群基金（R110324JG001）资助课题

Blind recognition of LDPC codes based on the convolutional neural network with cosine check relationship

Wenjie CHEN¹(), Pu ZHANG²^,*, Gaoxiang SHI³, Lin LIU², Xuan LIU²

1. The 10th Research Institute of China Electronics Technology Group Corporation，Chengdu 610036，China
2. Key Laboratory of Information Coding and Transmission，Southwest Jiaotong University，Chengdu 611756，China
3. Unit 93147 of the PLA，Chengdu 610041，China

Received:2024-08-05 Online:2025-09-25 Published:2025-09-16
Contact: Pu ZHANG E-mail:chenwenjiespring@163.com

摘要/Abstract

摘要：

针对低信噪比环境下低密度奇偶校验（low density parity check，LDPC）码的识别率低的问题，提出了一种基于余弦校验关系分布的卷积神经网络（convolutional neural network，CNN）算法。该算法基于码字与正确和错误校验矩阵的余弦校验关系统计分布间的差异性，利用LDPC码与候选集校验矩阵计算得到的余弦校验关系的统计特性作为CNN的输入，利用CNN的深层信息挖掘能力，设计一种结构简单的四层CNN模型，实现LDPC码的有效识别。仿真结果表明，仅使用一个码字的条件下，在信噪比为3.25 dB时，对码率1/2、2/3B、3/4A、3/4B、5/6，码长2304的LDPC码的正确识别率达到90%以上，与传统算法相比，性能提升了0.25~1.25 dB。

关键词: 低密度奇偶校验码, 闭集识别, 余弦校验关系, 卷积神经网络

Abstract:

Aiming at the problem of low recognition rate of low density parity check （LDPC） code in low signal-to-noise ratio environment, a convolutional neural network （CNN） algorithm based on the distribution of cosine check relations hip is proposed. This algorithm utilizes the statistical distribution differences of cosine check relations hip between codewords and correct and incorrect check matrices. The statistical characteristics of cosine check relations hip calculated from LDPC codes and candidate set check matrices are used as inputs for the CNN. By leveraging the deep information mining capability of CNN, a simple four-layer CNN model is designed to achieve effective recognition of LDPC codes. Simulation results show that with only one codeword, the correct recognition rate of LDPC codes with code rates of 1/2, 2/3B, 3/4A, 3/4B, and 5/6, and a code length of 2304, exceeds 90% at a signal-to-noise ratio of 3.25 dB. Compared to traditional algorithms, performance is improved by 0.25?1.25 dB.

Key words: low density parity check (LDPC) code, closed set recognition, cosine check relationship, convolutional neural network (CNN)

中图分类号:

TN 914.34

陈文洁, 张浦, 史高翔, 刘林, 刘烜. 基于余弦校验关系的卷积神经网络LDPC码盲识别[J]. 系统工程与电子技术, 2025, 47(9): 3117-3125.

Wenjie CHEN, Pu ZHANG, Gaoxiang SHI, Lin LIU, Xuan LIU. Blind recognition of LDPC codes based on the convolutional neural network with cosine check relationship[J]. Systems Engineering and Electronics, 2025, 47(9): 3117-3125.

图/表 10

图1

图2

图3

表1

图4

图5

图6

图7

图8

表2

参考文献 30

1	LONG F, TANG J Y, LI L F, et al. Adaptive modulation and coding with LDPC codes and retransmission for ultraviolet communication[J]. IEEE Photonics Journal, 2024, 16 (3): 7301610.
2	ZHANG Q W, LIU B Y, CHEN G W, et al. An improved adaptive coding and modulation scheme with hybrid switching standard for UAV-to-ground free space optical communication[J]. IEEE Photonics Journal, 2024, 16 (3): 7300108.
3	PENG H F, CAO C J, SUN Y, et al. Blind identification of channel codes under AWGN and fading conditions via deep learning[C]//Proc. of the International Conference on Networking and Network Applications, 2022: 67−73.
4	YANG J, YAN C Y, MA Y, et al. Recognition of punctured convolutional codes based on multi-scale CNN[C]//Proc. of the IEEE 98th Vehicular Technology Conference, 2023.
5	DUHOVNIKOV S, BALTACI A, GERA D, et al. Power consumption analysis of NB-IoT technology for low-power aircraft applications[C]//Proc. of the IEEE 5th World Forum on Internet of Things, 2019: 719−723.
6	WANG Y H, ZHANG W, SHI L N, et al. Blind recognition of RS codes based on channel condition determination[J]. IEEE Communications Letters, 2024, 28 (5): 1132- 1136. doi: 10.1109/LCOMM.2024.3372288
7	HONG S J, YUN W J, JEONG E R, et al. Blind detection of channel coding and interleaving using convolutional neural networks in tactical communications[C]//Proc. of the 15th International Conference on Ubiquitous and Future Networks, 2024: 560−562.
8	BAYESTEH A, YI E, NIKOPOUR H, et al. Blind detection of SCMA for uplink grant-free multiple-access[C]//Proc. of the 11th International Symposium on Wireless Communications Systems, 2014: 853−857.
9	LI L Q, HUANG Z P, ZHOU J, et al. Blind identification of LDPC codes based on a candidate set[J]. IEEE Communications Letters, 2021, 25 (9): 2786- 2789. doi: 10.1109/LCOMM.2021.3083921
10	WANG Y, CHE S L. Blind recognition of 5G LDPC codes over a candidate set[J]. IEEE Communications Letters, 2024, 28 (4): 744- 748. doi: 10.1109/LCOMM.2024.3359856
11	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. doi: 10.11959/j.issn.1000-436x.2017116
	YU P D, PENG H, GONG K X, et al. LDPC code reconstruction based on algorithm of finding low weight code-words[J]. Journal on Communications, 2017, 38 (6): 108- 117. doi: 10.11959/j.issn.1000-436x.2017116
12	刘洁, 刘凯. 基于深度学习的纠错编码方式识别[J]. 电子测量技术, 2019, 42 (16): 154- 158.
	LIU J, LIU K. Recognition of error correction coding based on deep learning[J]. Electronic Measurement Technology, 2019, 42 (16): 154- 158.
13	周靖. 信道编码识别分析方法研究[D]. 长沙: 国防科学技术大学, 2014.
	ZHOU J. Research of recognition and analyzation methods of channel codes[D]. Changsha: National University of Defense Technology, 2014.
14	YARDI A D, VIJAYAKUMARAN S, KUMAR A, et al. Blind reconstruction of binary cyclic codes from unsynchronized bitstream[J]. IEEE Trans. on Communications, 2016, 64 (7): 2693- 2706. doi: 10.1109/TCOMM.2016.2561931
15	GAUTAM N, LALL B. Blind channel coding identification of convolutional encoder and Reed-Solomon encoder using neural networks[C]//Proc. of the National Conference on Communications, 2020.
16	LIU Q, ZHANG H, YU P D, et al. Blind detection and identification of polar codes based on the parity check conformity[J]. IEEE Communications Letters, 2022, 26 (4): 728- 732. doi: 10.1109/LCOMM.2022.3149797
17	陈泽亮, 彭华, 巩克现, 等. 误码条件下LDPC码参数的盲估计[J]. 电子学报, 2018, 46 (3): 652- 658. doi: 10.3969/j.issn.0372-2112.2018.03.021
	CHEN Z L, PENG H, GONG K X, et al. A method for blind recognition of LDPC codes in a noisy environment[J]. Acta Electronica Sinica, 2018, 46 (3): 652- 658. doi: 10.3969/j.issn.0372-2112.2018.03.021
18	XIA T, WU H C. Novel blind identification of LDPC codes using average LLR of syndrome a posteriori probability[C]//Proc. of the 12th International Conference on ITS Telecommunications, 2012: 12−16.
19	袁磊, 杨艳娟, 郭毅, 等. 相关噪声下基于深度学习的LDPC码码率半盲识别算法[J]. 系统工程与电子技术, 2025, 47 (4): 1335- 1345.
	YUAN L, YANG Y L, GUO Y, et al. Deep learning-based semi-blind identification algorithm for code rates of LDPC codes over correlated noise[J]. Systems Engineering and Electronics, 2025, 47 (4): 1335- 1345.
20	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.
	TAO Z Y, YAN M H, LIU Y. Channel coding closed set recognition based on temporal convolutional network[J]. Journal of Huazhong University of Science and Technology （Natural Science Edition）, 2022, 50 (3): 12- 17.
21	LI X Y, ZHANG Y F, WANG S J, et al. A maximum generating function method for blind recognition of LDPC codes[C]//Proc. of the 4th International Conference on Computer Communication and Artificial Intelligence, 2024: 378−382.
22	MOOSAVI R, LARSSON E G. A fast scheme for blind identification of channel codes[C]//Proc. of the IEEE Global Telecommunications Conference, 2011.
23	MOOSAVI R, LARSSON E G. Fast blind recognition of channel codes[J]. IEEE Trans. on Communications, 2014, 62 (5): 1393- 1405. doi: 10.1109/TCOMM.2014.050614.130297
24	魏刚, 张云冲, 戴旭初. 一种改进的信道编码盲识别算法[J]. 遥测遥控, 2016, 37 (2): 13- 20. doi: 10.3969/j.issn.2095-1000.2016.02.003
	WEI G, ZHANG Y C, DAI X C. A modified algorithm for blind recognition of channel codes[J]. Journal of Telemetry, Tracking and Command, 2016, 37 (2): 13- 20. doi: 10.3969/j.issn.2095-1000.2016.02.003
25	WU Z J, ZHANG L M, ZHONG Z G, et al. Blind recognition of LDPC codes over candidate set[J]. IEEE Communications Letters, 2020, 24(1): 11−14.
26	刘仁鑫, 张立民, 钟兆根, 等. 基于最大余弦比的LDPC码闭集识别[J]. 信号处理, 2020, 36 (8): 1234- 1242.
	LIU R X, ZHANG L M, ZHONG Z G, et al. Closed set identification of LDPC codes based on maximum cosine ratio[J]. Journal of Signal Processing, 2020, 36 (8): 1234- 1242.
27	LI L Q, HUANG Z P, ZHOU J. A two-stage maximum a posterior probability method for blind identification of LDPC codes[J]. IEEE Signal Processing Letters, 2021, 28, 111- 115. doi: 10.1109/LSP.2020.3047334
28	ZOU L Q, WU H L, LIU R, et al. A new method for LDPC blind recognition over a candidate set using Kullback-Leibler divergence[J], IEEE Communications Letters, 2024, 28(5): 964−968.
29	LI L Q, XIE L H, HUANG Z P, et al. A deep learning based method for blind recognition of LDPC codes[C]//Proc. of the 2nd Information Communication Technologies Conference, 2021: 112−116.
30	杨宗方, 张天骐, 马焜然, 等. 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.
	YANG Z F, ZHANG T Q, MA K R, et al. Blind identification of channel coding types based on deep neural networks[J]. Systems Engineering and Electronics, 2024, 46 (5): 1820- 1829.

类型	超参数
优化器	Adam
学习率	0.001
批量大小	128
激活函数	Relu
迭代次数	20
填充方式	same

算法	时间
基于SPP最大值算法	7.81
基于PPLLR均值最大值算法	7.11
基于最大余弦比算法	7.37
基于KL散度最大值算法	4.72
本文CNN算法	3.96

[1]	蒋明煜, 张顺生, 肖思瑶. 面向轻量级交叉注意力卷积网络的SAR目标识别[J]. 系统工程与电子技术, 2025, 47(9): 2853-2861.
[2]	段阿敏, 张朝辉. 基于二次分解的混合神经网络蜂窝流量预测[J]. 系统工程与电子技术, 2025, 47(5): 1687-1697.
[3]	孙志国, 王一珂, 宁晓燕. LDPC的分段多因子最小和译码算法[J]. 系统工程与电子技术, 2025, 47(5): 1698-1705.
[4]	袁磊, 杨艳娟, 郭毅, 戴鹏. 相关噪声下基于深度学习的LDPC码码率半盲识别算法[J]. 系统工程与电子技术, 2025, 47(4): 1335-1345.
[5]	付卫红, 张鑫钰, 刘乃安. 基于多尺度融合神经网络的同频同调制单通道盲源分离算法[J]. 系统工程与电子技术, 2025, 47(2): 641-649.
[6]	邵永琪, 杨丽花, 常澳, 任露露. RIS辅助的OFDM系统中时变信道估计方法[J]. 系统工程与电子技术, 2025, 47(1): 324-331.
[7]	于蕾, 邓秋月, 郑丽颖, 吴昊宇. 二阶逐层特征融合网络的图像超分辨重建[J]. 系统工程与电子技术, 2024, 46(2): 391-400.
[8]	郑仁乐, 李东阳, 刘文学, 万金涛, 刘学勇, 李金海. LDPC码的分层自适应最小和译码算法[J]. 系统工程与电子技术, 2024, 46(12): 4231-4237.
[9]	肖博文, 马泽远, 鲁天宇, 夏群利. 平台式导引头隔离度模型在线辨识方法[J]. 系统工程与电子技术, 2024, 46(11): 3595-3604.
[10]	季然, 肖茂森, 李硕, 刘宇, 罗湛仪, 程嘉维. 基于机器学习的MRTD客观测试方法研究[J]. 系统工程与电子技术, 2024, 46(10): 3265-3270.
[11]	杨帆, 马萍, 李伟, 杨明. 基于孪生网络的仿真模型智能排序评估方法[J]. 系统工程与电子技术, 2023, 45(7): 2060-2068.
[12]	姜雨, 袁琪, 胡志韬, 吴薇薇, 顾欣. 基于气象因素的机场进离港延误预测[J]. 系统工程与电子技术, 2023, 45(6): 1722-1731.
[13]	汪锐, 张天骐, 安泽亮, 王雪怡, 方竹. 基于联合特征参数和一维CNN的MIMO-OFDM系统调制识别算法[J]. 系统工程与电子技术, 2023, 45(3): 902-912.
[14]	闫啸家, 梁伟阁, 张钢, 佘博, 田福庆. 基于RCNN-ABiLSTM的机械设备剩余寿命预测方法[J]. 系统工程与电子技术, 2023, 45(3): 931-940.
[15]	宋波涛, 许广亮. 基于LSTM与1DCNN的导弹轨迹预测方法[J]. 系统工程与电子技术, 2023, 45(2): 504-512.

基于余弦校验关系的卷积神经网络LDPC码盲识别

Blind recognition of LDPC codes based on the convolutional neural network with cosine check relationship

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 30

相关文章 15

编辑推荐

Metrics

本文评价