Home > Published Issues > 2024 > Volume 19, No. 5, 2024 >
JCM 2024 Vol.19(5): 229-241
Doi: 10.12720/jcm.19.5.229-241

Attention Turbo-Autoencoder for Improved Channel Coding and Reconstruction

Kayode A. Olaniyi1, R. Heymann1, and Theo G. Swart1,2*
1.Department of Electrical and Electronic Engineering Science, University of Johannesburg, South Africa
2.Center for Telecommunications, University of Johannesburg, South Africa
Email: 217097911@student.uj.ac.za (K.A.O.); reolyn.heymann@gmail.com (R.H.); tgswart@uj.ac.za (T.G.S.)
*Corresponding author

Manuscript received November 8, 2023; revised January 21, 2024; accepted February 6, 2024; published May 8, 2024.

Abstract—Channel coding and signal reconstruction are critical tasks in communication systems to ensure reliable and efficient data transmission. Traditional approaches, such as turbo codes and Low-Density Parity-Check (LDPC) codes, have been widely used for these tasks. However, these methods may suffer from limitations in accurately capturing the complex channel characteristics and effectively handling noisy environments. To address these challenges, this research proposes deep learning-based channel codes that integrate an attention mechanism into turbo code autoencoders, referred to as ATT-TurboAE, to enhance channel coding and reconstruction performance. The attention mechanism selectively focuses on informative features while suppressing noise and interference, improving the accuracy and robustness of the system. The proposed approach is evaluated using simulated datasets and compared with traditional turbo autoencoders. The results demonstrate that the attention mechanism significantly improves the performance of channel coding and signal reconstruction, achieving higher accuracy and better noise resilience. This research contributes to the advancement of communication systems by introducing a novel technique that enhances turbo autoencoders through the incorporation of attention mechanisms, leading to improved channel coding and reconstruction performance.

Keywords—channel coding, TurboAE, attention mechanism, encoder-decoder, channel matrix, autoencoder

Cite: Kayode A. Olaniyi, R. Heymann, and Theo G. Swart, “Attention Turbo-Autoencoder for Improved Channel Coding and Reconstruction," Journal of Communications, vol. 19, no. 5, pp. 229-241, 2024.

Copyright © 2024 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.