End-To-End Anti-counterfeiting Information Coding System Based on MLP-LSTM Joint Optimization

LI Siqi; CAO Peng

doi:10.19554/j.cnki.1001-3563.2025.23.023

PDF(719 KB)

Packaging Engineering ›› 2025, Vol. 46 ›› Issue (23) : 212-220. DOI: 10.19554/j.cnki.1001-3563.2025.23.023

Automatic and Intelligent Technology

End-To-End Anti-counterfeiting Information Coding System Based on MLP-LSTM Joint Optimization

LI Siqi, CAO Peng^*

Author information +

History +

Abstract

To address the problems of existing Printed Quantum Dot anti-counterfeiting algorithms relying on experience design, leading to low efficiency and difficulty in improving performance, the work aims to propose an end-to-end neural-network-based encoding and decoding system for this scenario. By incorporating the self-supervised optimization mechanism of neural-networks, this system can encode the anti-counterfeiting information efficiently, thereby improving the design efficiency and error detection or correction capabilities of the encoding and decoding system. First, a Multilayer Perceptron neural network was used as a dot matrix data generator for printing quantum dots. The Hamming distance regularities were introduced to the loss function as constraint, thereby increasing the difference between the generated anti-counterfeiting codes and improve the error detection and correction tolerance. For damaged anti-counterfeiting code words, a Long Short-Term Memory (LSTM) neural network was used for sequential decoding. Finally, by jointly optimizing the network parameters of the encoder and decoder, an integrated encoding and decoding framework for printed anti-counterfeiting information was achieved. Through simulated noise attacks, the bit error rates (BER) were compared within a signal-to-noise ratio (SNR) range of 0~8 dB. Experimental results showed that the LSTM decoder outperformed MLP and CNN decoders in error detection and correction. Under a 6 dB noise attack, the LSTM decoder achieved a BER (5.19×10^-4), lower than those of the MLP decoder (2.23×10^-3) and CNN decoder (9.02×10^-4). The proposed end-to-end printed quantum dot anti-counterfeiting information encoding and decoding system can autonomously generate anti-counterfeiting code words with strong error detection and correction capabilities, effectively overcoming the limitations of traditional methods in terms of designing inefficiency and difficulty in iterative performance improvement.

Key words

printed quantum dots / printing anti-counterfeiting / auto encoder and decoder / Multilayer Perceptron (MLP) / Long Short-Term Memory (LSTM)

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

LI Siqi, CAO Peng. End-To-End Anti-counterfeiting Information Coding System Based on MLP-LSTM Joint Optimization[J]. Packaging Engineering. 2025, 46(23): 212-220 https://doi.org/10.19554/j.cnki.1001-3563.2025.23.023

References

[1] 赵文康, 曹鹏. 基于印刷量子点的可靠性复合光谱图像编解码算法研究[J]. 包装工程, 2025, 46(7): 173-182.
ZHAO W K, CAO P.Encoding and Decoding Algorithms for Image Information of Composite Spectrum Based on Printed Quantum Dots[J]. Packaging Engineering, 2025, 46(7): 173-182.
[2] 邱英英, 曹鹏. 印刷量子点点阵信息可靠性编解码算法[J]. 计算机科学与应用, 2023, 13(3): 617-625.
QIU Y Y, CAO P.Reliability Coding and Decoding Algorithm of Printed Quantum Dot Lattice Information[J]. Computer Science and Application, 2023, 13(3): 617-625.
[3] 王育军, 曹鹏, 王明飞, 等. 基于Turbo码的印刷量子点信息隐藏算法研究[J]. 数字印刷, 2022(4): 195-206.
WANG Y J, CAO P, WANG M F, et al.Research on Information Hiding Algorithm for Printing-Quantum-Dots Based on Turbo Codes[J]. Digital Printing, 2022(4): 195-206.
[4] 魏超. 基于混沌加密的点阵防伪码研究及应用[D]. 哈尔滨: 哈尔滨工业大学, 2017: 4.
WEI C.Research and Application of Dot Matrix Anti-Counterfeiting Code Based on Chaotic Encryption[D]. Harbin: Harbin Institute of Technology, 2017: 4.
[5] WANG X A, WICKER S B.An Artificial Neural Net Viterbi Decoder[J]. IEEE Transactions on Communications, 1996, 44(2): 165-171.
[6] GRUBER T, CAMMERER S, HOYDIS J, et al.On Deep Learning Based Channel Decoding[C]// In 2017 51st Annual Conference on Information Sciences and Systems (CISS), 2017: 1-6.
[7] LYU W, ZHANG Z Y, JIAO C X, et al.Performance Evaluation of Channel Decoding with Deep Neural Networks[C]// 2018 IEEE International Conference on Communications (ICC). Kansas City, MO. IEEE, 2018: 1-6.
[8] LIANG F S, LU S, UENG Y L.Deep-Learning-Aided Successive Cancellation List Flip Decoding for Polar Codes[J]. IEEE Transactions on Cognitive Communications and Networking, 2024, 10(2): 374-386.
[9] LI J, ZHOU L J, LI Z Q, et al.Deep Learning-Assisted Dynamic-SCLF Decoding of
Polar Codes[J]. IEEE Transactions on Cognitive Communications and Networking, 2024, 10(3): 836-851.
[10] O'SHEA T, HOYDIS J. An Introduction to Deep Learning for the Physical Layer[J]. IEEE Transactions on Cognitive Communications and Networking, 2017, 3(4): 563-575.
[11] DÖRNER S, CAMMERER S, HOYDIS J, et al. Deep Learning Based Communication over the Air[J]. IEEE Journal of Selected Topics in Signal Processing, 2018, 12(1): 132-143.
[12] JIANG Y H, KIM H, ASNANI H, et al. Turbo Autoencoder: Deep Learning Based Channel Codes for Point-to-Point Communication Channels[EB/OL].2019: arXiv: 1911.03038. https://arxiv.org/abs/1911.03038
[13] 胡启蕾, 许佳龙, 李伦, 等. 信噪比自适应Turbo自编码器信道编译码技术[J]. 无线电通信技术, 2022, 48(4): 680-688.
HU Q L, XU J L, LI L, et al.Turbo Autoencoder Adapting to Signal-to-Noise Ratio for Channel Coding and Decoding[J]. Radio Communications Technology, 2022, 48(4): 680-688.
[14] 金林贤, 王旭东, 吴楠. Neural-Polar码: 一种基于深度学习的新型信道编码方案[J]. 重庆邮电大学学报(自然科学版), 2024, 36(3): 430-437.
JIN L X, WANG X D, WU N.Neural-Polar Code: An Inventing Channel Coding Scheme Based on Deep Learning[J]. Journal of Chongqing University of Posts and Telecommunications (Natural Science Edition), 2024, 36(3): 430-437.
[15] 曹明华, 王瑞, 张悦, 等. CNN-AE在超奈奎斯特无线光通信端到端系统中的性能[J]. 重庆邮电大学学报(自然科学版), 2024, 36(1): 181-190.
CAO M H, WANG R, ZHANG Y, et al.End-to-End Performance of a CNN-AE Based Faster-than-Nyquist Rate Free Space Optical Communication System[J]. Journal of Chongqing University of Posts and Telecommunications (Natural Science Edition), 2024, 36(1): 181-190.
[16] 宋旭冉. 基于学习的极化码译码算法研究与应用[D]. 杭州: 浙江大学, 2020: 3-4.
SONG X R.Research and Application on Decoding Algorithms of Polar Codes With Deep Learning[D]. Hangzhou: Zhejiang University, 2020: 3-4.
[17] 曹鹏, 刘喆灿, 衣旭梅. 半色调加网与信息隐藏技术[M]. 北京: 电子工业出版社, 2013: 113-130.
CAO P, LIU Z C, YI X M.Halftone Screening and Information Hiding Technology[M]. Beijing: Publishing House of Electronics Industry, 2013: 113-130.
[18] KUZNETSOV Y V.High-Definition Halftone Printing[M]. Principles of Image Printing Technology. Cham: Springer International Publishing, 2021: 331-352.
[19] ULICHNEY R.Digital Halftoning[M]. Cambridge: MIT Press, 1987.
[20] HOCHREITER S, SCHMIDHUBER J.Long Short-Term Memory[J]. Neural Computation, 1997, 9(8): 1735-1780.
[21] HU J, CUI J W, XIANG L P, et al.End-to-End Design of Polar Coded Integrated Data and Energy Networking[J]. IEEE Transactions on Communications, 2024, 72(11): 7017-7031.
[22] GRAVES A, MOHAMED A R, HINTON G.Speech Recognition with Deep Recurrent Neural Networks[C]// 2013 IEEE International Conference on Acoustics, Speech and Signal Processing. Vancouver, BC, Canada. IEEE, 2013: 6645-6649.
[23] ARIKAN E.Channel Polarization: A Method for Constructing Capacity-Achieving Codes for Symmetric Binary-Input Memory Channels[J]. IEEE Transactions on Information Theory, 2009, 55(7): 3051-3073.
[24] 王旭东, 林彬, 张凯尧, 等. 一种改进的CNN端到端自编码器通信系统[J]. 电讯技术, 2020(2): 147-152.
WANG X D, LIN B, ZHANG K Y, et al.An Improved CNN End-to-End Self-Encoder Communication System[J]. Telecommunications Technology, 2020(2): 147-152.