Abstract
We present a convolutional neural network (CNN) for the decoding of a terminated convolutional code (CC). For this use cases, an unlimited amount of labeled training data can be generated. However, the number of code words, i.e., pattern, to be learned by the CNN increases exponentially with the dimension of the code. Therefore, scalability of the neural network is of critical importance. The CNN is trained with a CC with small code word length, i.e., a limited number of code words. Therefore the training complexity is feasable. As the CNN is properly matched to dimension and structural properties of the CC, it can actually learn the structure of the CC. This allows to upscale the CNN decoder to match to CCs with larger code dimension. A retraining of weights is not required. The upscaled CNN successfully decodes code words it has never seen during the training process (generalization).
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
O’Shea, T., Hoydis, J.: An introduction to deep learning for the physical layer. IEEE Trans. Cognitive Commun. Networking 3(4) (2017)
Farsad, N., Goldsmith, A.: Detection Algorithms for Communication Systems using Deep Learning, arXiv preprint arXiv:1705.08044 (2017)
Gruber, T., Cammerer, S., Hoydis, J., ten Brink, S.: On deep learning-based channel decoding. In: 51st Annual Conference on Information Sciences and Systems (CISS), Baltimore (2017)
Teich, W.G., Liu, R., Belagiannis, V.: Deep learning versus high-order recurrent neural network based decoding for convolutional codes. In: GLOBECOM 2020–2020 IEEE Global Communications Conference, pp. 1–7 (2020). https://doi.org/10.1109/GLOBECOM42002.2020.9348117
Cammerer, S., Gruber, T., Hoydis, J., ten Brink, S.: Scaling deep learning-based decoding of polar codes via partitioning. In: GLOBECOM 2017–2017 IEEE Global Communications Conference, Singapore (2017)
Caid, W.R., Means, R.W.: Neural network error correcting decoders for block and convolutional codes. In: GLOBECOM 1990, San Diego, CA 2, pp. 1028–1031 (1990)
Marcone, G., Zincolini, E., Orlandi, G.: An efficient neural decoder for convolutional codes. Eur. Trans. Telecommun. 6, 439–445 (1995)
Teich, W.G., Boettrich, M.: Optimal Decoding of Convolutional Codes using a Multilayer Perzeptron, ITG-Fachtagung Codierung für Quelle, Kanal und Übertragung, München, ITG-Fachbericht 130, pp. 73–80. VDE Verlag GmbH, Berlin (1994)
Teich, W.G., Lindner, J.: A novel decoder structure for convolutional codes based on a multilayer perceptron. In: IEEE ICNN’95, Perth, Western Australia 1, pp. 449–454 (1995)
Teich, W.G.: Iterative decoding of one-dimensional convolutional codes. In: 4th International Symposium on Communication Theory and Applications, pp. 52–53. Ambleside, UK (1997)
Mostafa, M., Teich, W.G., Lindner, J.: Analog realizatiuon of iterative threshold decoding based on high-order recurrent neural networks. In: International Conference on Signal Processing and Communication Systems (ICSPCS), Gold Coast, Australia (2010)
Mostafa, M., Teich, W.G., Lindner, J.: Analysis of high order recurrent neural networks for analog decoding. In: 7th International Symposium on Turbo Codes and Iterative Information Processsing. ISTC), Goeteborg, Sweden (2012)
Teich, W.G., Teich, H., Oliveri, G.: From iterative threshold decoding to a low-power high-speed analog VLSI decoder implementation. In: Rojas, I., Joya, G., Catala, A. (eds.) IWANN 2019. LNCS, vol. 11507, pp. 615–628. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-20518-8_51
Zhang, Z., Yao, D., Xiong, L., Ai, B., Guo, S.: A convolutional neural network decoder for convolutional codes. In: Gao, H., Feng, Z., Yu, J., Wu, J. (eds.) ChinaCom 2019. LNICST, vol. 313, pp. 113–125. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-41117-6_10
Johnson, S.: Iterative Error Correction: Turbo. Cambridge University Press, Low-Density Parity-Check and Repeat-Accumulate Codes (2010)
Hagenauer, J., Offer, E., Méasson, C., Mörz, M.: Decoding and equalization with analog non-linear networks. Eur. Trans. Telecommun. 10, 659–680 (1999)
Goodfellow, I., Bengio, Y., Courville, A.: Deep Learning. MIT Press (2016)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Teich, W.G., Pan, W. (2023). Scalable Convolutional Neural Networks for Decoding of Terminated Convolutional Codes. In: Rojas, I., Joya, G., Catala, A. (eds) Advances in Computational Intelligence. IWANN 2023. Lecture Notes in Computer Science, vol 14134. Springer, Cham. https://doi.org/10.1007/978-3-031-43085-5_4
Download citation
DOI: https://doi.org/10.1007/978-3-031-43085-5_4
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-43084-8
Online ISBN: 978-3-031-43085-5
eBook Packages: Computer ScienceComputer Science (R0)