Сообщение

Simulation of Machine Learning Methods to Improve Communication Quality in a Short-Wave Radio Channel

 
 orcid Mikhail Isakov,  orcid Olga Simonina

The Bonch-Bruevich Saint Petersburg State University of Telecommunications,
St. Petersburg, 193232, Russian Federation

DOI 10.31854/2307-1303-2025-13-4-54-70

EDN NASEQO

 Full text

XML JATS

Abstract

Relevance. In the current conditions of telecommunications development, the shortwave band retains its significance as a cost-effective solution for long-range communication. At the same time, the existing methods of signal processing in HF channels require improvement in efficiency in the conditions of a complex ionospheric environment. The use of machine learning technologies opens up new opportunities for improving the quality of communication. The purpose of the study is to improve the quality of communication in the short-wave (SW) channel by applying machine learning methods for predicting ionospheric parameters and demodulating signals using OFDM technologies. The research methods include the simulation of LSTM-networks for predicting ionospheric parameters, as well as the application of convolutional (CNN) and multilayer perceptrons (MLP) for demodulating signals. The work uses quality metrics such as RMSE and BER, as well as ITU-R P.533-14 recommendations for modeling signal propagation conditions.The scientific novelty lies in a comprehensive approach to improving communication quality in the HF channel, which combines ionospheric parameter forecasting with machine learning techniques for signal demodulation. A comparative evaluation of the effectiveness of various neural network architectures in the HF channel is proposed. The results of the study showed that the use of CNN-demodulators provides the best signal reception quality at low SNR values, demonstrating a gain of up to 2.5 dB compared to the classical correlation method. LSTM-networks showed high efficiency in predicting the maximum applicable frequencies and other ionospheric parameters. The practical significance of this work is to develop methods for improving the quality of communication in the HF band, which can be used to create adaptive radio communication systems with automatic selection of operating frequencies and modulation parameters.

Keywords

short-wave communication, OFDM, machine learning, LSTM-networks, CNN, MLP, demodulation, ionosphere forecasting, communication quality

Reference for citation

Isakov M., Simonina O. Simulation of Machine Learning Methods to Improve Communication Quality in a Short-Wave Radio Channel // Telecom IT. 2025. Vol. 13. Iss. 4. PP. 54‒70. (in Russian). DOI: 10.31854/2307-1303-2025-13-4-54-70. EDN: NASEQO
References

1. Vatrukhin E. M. New Opportunities of Using Short-Wave Radio Communication When the Aerospace Defense Tasks Are Solved by Combat Aviation // Journal of "Almaz-Antey" Air and Space Defence Corporation. 2017. Iss. 2 (21). PP. 16-20. (in Russian) EDN: YOBVLS

2. Chikhachev A. V., Budko P. A., Schmidt A. A. Application of Machine Learning Algorithms to Solve Problems of Predicting the Technical Condition of Radio Communication Facilities // Telecommunications and Communications. 2024. Iss. 3 (3). PP. 33-40. (in Russian) DOI: 10.24682/3034-4050-2024-3-33-40. EDN: IDTCNI

3. Shirokov S. Yu. Research on Influence of Artificial Intelligence for Optimization Data Transfer // Science Bulletin. 2024. Vol. 4. Iss. 12 (81). PP. 1685-1689. (in Russian) EDN: AREGOJ

4. Egorov V. V., Mingalev A. N., Shcheglova E. F. Multivariable Adaptation of HF Data Transmission Systems with OFDM Signals // Means of Communication Equipment. 2021. Iss. 3 (155). PP. 18-28. (in Russian) EDN: LYRHQF

5. Zemlyanov I. S. Modems with Orthogonal Subcarriers for Mobile Shortwave Communication Systems with Adaptation to Radio Wave Propagation Conditions. Ph. D. Thesis. Omsk, 2016. 168 p. (in Russian) EDN: NQMQRT

6. Konkin N. A. Methodology and Algorithm for Determining the Periods of Operational Forecasting of the Maximum Usable Short Wave Communication Frequencies Based on the XGBoost Machine Learning Algorithm // Vestnik of Volga State University of Technology. Series: Radio Engineering and Infocommunication Systems. 2022. Iss. 3 (55). PP. 6-16. (in Russian) DOI: 10.25686/2306-2819.2022.3.6. EDN: ZPKQBM

7. Wang J., Shi Y., Yang C., Feng F. A Review and Prospects of Operational Frequency Selecting Techniques for HF Radio Communication // Advances in Space Research. 2022. Vol. 69. Iss. 8. PP. 2989-2999. DOI: 10.1016/j.asr.2022.01.026. EDN: GAQWAW

8. Liu X., Xu Yu., Cheng Yu., Li Ya., Zhao L. et al. A Heterogeneous Infor-mation Fusion Deep Reinforcement Learning for Intelligent Frequency Selection of HF Communication // China Communications. 2018. Vol. 15. Iss. 9. PP. 73-84. DOI: 10.1109/CC.2018.8456453

9. Oyedare T., Shah V. K., Jakubisin D. J., Reed J. H. Interference Suppression Using Deep Learning: Current Approaches and Open Challenges // IEEE Access. 2022. Vol. 10. PP. 66238-66266. DOI: 10.1109/access.2022.3185124. EDN: VPQSLS

10. Liu Ch., Chen Yu., Yang Sh. H. Deep Learning Based Detection for Communications Systems with Radar Interference // IEEE Transactions on Vehicular Technology. 2022. Vol. 71. Iss. 6. PP. 6245-6254. DOI: 10.1109/tvt.2022.3158692. EDN: OECFXQ

11. Solovyeva E. B., Zubarev A. V. Neural Model of Nonlinear Signal Distortion Compensator for Digital Communication Channel // Journal of the Russian Universities. Radioelectronics. 2013. Iss. 4. PP. 30-34. (in Russian) EDN: RUXAHR

12. Malygin I., Belkov S., Tarasov A., Usvyatsov M. Machine Learning Methods in Classification of Radio Signals // Trudy MAI. 2017. Iss. 96. P. 15. (in Russian) EDN: ZWUHFP

13. Rapakov G. G., Gorbunov V. A., Dianov S. V., Elizarova L. V. Research of the LSTM Neural Network Approach in Time Series Modeling // Cherepovets State University Bulletin. 2023. Iss. 3 (114). PP. 47-54. (in Russian) DOI: 10.23859/1994-0637-2023-3-114-4. EDN: AEWHEN

14. Ji S., He G., Yu Q., Shi Ya., Hu Ju., et al. A Short-Term Forecasting Method for High-Frequency Broadcast MUF Based on LSTM // Atmosphere. 2024. Vol. 15. Iss. 5. P. 569. DOI: 10.3390/atmos15050569. EDN: PSQPAQ

15. Shenvi N., Virani H. Forecasting of Ionospheric Total Electron Content Data Using Multivariate Deep LSTM Model for Different Latitudes and Solar Activity // Journal of Electrical and Computer Engineering. 2023. Vol. 2023. P. 2855762. DOI: 10.1155/2023/2855762. EDN: XUUDIO

16. Xiong P., Zhang X., Zhai D., Long C., Zhou H., et al. Long Short-Term Memory Neural Network for Ionospheric Total Electron Content Forecasting over China // Space Weather. 2021. Vol. 19. Iss. 4. P. e2020SW002706. DOI: 10.1029/2020SW002706. EDN: QHEYCE

17. Ivanov M. S., Lenshin A. V. Statistical Testing of Methods of Reception and Demodulation of Signals with Compensation of Non-Orthogonal Simulated Noise // Vestnik of Voronezh Institute of the Ministry of Interior of Russia. 2023. Iss. 4. PP. 149-158. (in Russian) EDN: JRLYZZ

18. Eliseev S. N., Filimonova L. N. The Effect of Simultaneous Effects of Fast Fading and Frequency Shift in the Radio Channel on the OFDM Signal // The V Scientific Forum "Telecommunications: Theory and Technology" (TTT-2021). Proceedings of the XXIII International Scientific and Technical Conference "Problems of Telecommunications Engineering and Technology" (PTETT-2021, 23-26 November 2021, Samara, Russia). 2021. PP. 81-82. (in Russian) EDN: GAVLDS

19. Hassan H. A., Mohamed M. A., Essai M. H., Mubarak A. S., Esmaiel H., et al. An Efficient and Reliable OFDM Channel State Estimator Using Deep Learning Convolutional Neural Networks // Journal of Engineering Sciences. 2023. Vol. 51. Iss. 6. PP. 32-48. DOI: 10.21608/jesaun.2023.215113.1236. EDN: JCUMCA

20. Lobov E. M., Alaa A. Review of Existing Methods for Correcting Inter-Symbol Distortions of Radio Signals in Digital Communication Systems Using Machine Learning // Telecommunications and Information Technologies. 2023. Vol. 10. Iss. 1. PP. 109-119. EDN: MQABZM (in Russian)

21. Zhao R., Wang J., Li J. An End-to-End Demodulation System Based on Convolutional Neural Networks // Journal of Physics: Conference Series. 2nd International Conference on Computer Science and Communication Technology (ICCSCT, 29-31 July 2021, Beijing, China). 2021. Vol. 2026. Iss. 1. P. 012006. DOI: 10.1088/1742-6596/2026/1/012006

22. Maranhão J. P. A., da Costa J. P. C. L., de Freitaset E. P., Javidial E. Noise-Robust Multilayer Perceptron Architecture for Distributed Denial of Service Attack Detection // IEEE Communications Letters. 2020. Vol. 25. Iss. 2. PP. 402-406. DOI: 10.1109/LCOMM.2020.3032170

23. Li J., Zhang M., Xu K., Dickerson J., Ba J. How Does a Neural Network's Architecture Impact Its Robustness to Noisy Labels? // Advances in Neural Information Processing Systems. 2021. Vol. 34. DOI: 10.48550/arXiv.2012.12896

24. Balamurugan S. P. A Comprehensive Study on MLP and CNN, and the implementation of Multi-Class Image Classification Using Deep CNN // Machine Learning and Deep Learning Techniques for Medical Science. CRC Press, 2022. PP. 1-25. DOI: 10.1201/9781003217497-1

 

cc-by This article is distributed under a license Creative Commons Attribution 4.0 License.

cc0  The metadata of the article is distributed under a license CC0 1.0 Universal


 
 
войти

Авторизация