Methodology for estimating cyclostationarity intervals of the probabilistic component of dynamic risks in aerial communication networks using combined SCD-analysis and LSTM-correction


Аuthors

Ananjev A. V.1, 2*, Pechkarev V. A.2**, Moiseev S. I.3

1. Plekhanov Russian University of Economics, PRUE, 36, Stremyanny per., Moscow, 117997, Russia
2. Air force academy named after professor N.E. Zhukovskogo and Y. A. Gagarin, 54a Starye Bolshevikov str., Voronezh, 394064, Voronezh Region
3. Voronezh State Technical University, VSTU, 14, Moskovsky prospect, Voronezh, 394026, Russia

*e-mail: Ananyev-Alexandr@yandex.ru
**e-mail: Val_pechkarev@mail.ru

Abstract

The paper proposes a methodology for the discrete risk assessment based on a hybrid method for estimating cyclostationarity intervals. The methodology differs from existing approaches by combining spectral correlation analysis (SCD) with a lightweight recurrent neural network featuring long short‑term memory (LSTM). The methodology enables justification of the requirements for the communication system with respect to addressing the problem of signal non‑stationarity, which stems from the dynamic nature of the changing environment typical of airborne communication networks. In the proposed methodology, SCD analysis provides an initial assessment of the signals’ cyclostationary properties, while the LSTM model adaptively adjusts the intervals taking into account changes in network parameters. A key aspect of the proposed hybrid SCD‑LSTM approach is the adaptation of the neural network model to the high dynamics of changes in airborne communication networks.
Keywords: overhead communication networks; dynamic risks; cyclostationarity; statistical sampling; risk assessment.

Keywords:

overhead communication networks; dynamic risks; cyclostationarity; statistical sampling; risk assessment

References

  1. Petrov I.V., Sidorov A.S. Application of lightweight neural networks in complex computer tasks // "Artificial Intelligence and Technology". 2022. № 5(12). С. 45-52.
  2. Ananjev A.V., Erzin I.Kh., Filatov S.V., Shcherbakov A.A. Aeromobile Communication Network – An Efficient Relay System for the Airborne Layer of the Integrated Automated Digital Communication System in Armed Conflict Conditions // Military Thought. 2017. No. 4. P. 26-34.
  3. Risk // The Great Russian Encyclopedia: [Electronic Resource]. URL: https://old.bigenc.ru/economics/text/3510546 (Date of Access: 02.05.2025).
  4. Ananjev A.V., Pechkarev V.A. Algorithm for Evaluating Feature Coefficients of Loss under Information Transmission Risks in Airborne Communication Networks // Proceedings of the Moscow Aviation Institute. 2025. No. 141. URL: https://trudymai.ru/published.php?ID=184511.
  5. Goodfellow, Y. Deep Learning / Y. Goodfellow, Y. Bengio, A. Courville; translated from English by A. A. Slonkina. – Moscow: DMK Press, 2022. 652 p.
  6. Petrov, A.V.; Smirnov, I.K. Noise immunity analysis of the wireless communication networks based on UAVs // Proc. of MAI. 2021. №112. URL: www.mai.ru/science/trudy (date of address 24.04.2025).
  7. Le Zheng, Marco Lops, Yonina C. Eldar, Xiaodong Wang. Radar and Communication Co-existence: an Overview [Electronic resource]. 2019. URL: https://arxiv.org/pdf/1902.08676.pdf (accessed 20.05.2025).
  8. Lebedev, S.M.; Fedorov, R.A. Methods of the non-stationary signals processing under the dynamic interference condition // Proc. of MAI. 2022. №115. URL: www.mai.ru/science/trudy (date of address 24.04.2025).
  9. Goncharenko, V.I.; Kletskov, D.A.; Serkin, F.B.; Shevtsov, V.A. Use of the cyclostationarity properties of the information radio signals for the blind detection of the high-speed wireless channels of the information transmission // Control, communication and security systems. 2023. №1. С. 110-124.
  10. Efimov, E.N.; Shevgunov, T.Y.; Zhukov, D.M. Analysis of the cyclostationary radio signals using Realtek RTL2832U and Qt framework // Elektrosvyaz. 2018. №12. С. 43-47.
  11. Gardner W.A. Cyclostationarity in Communications and Signal Processing. New York: IEEE Press, 1994. 458 p. ISBN 0-7803-1148-9.
  12. Grishin, V.P.; Orlov, A.S. Adaptive routing algorithms in airborne communication networks // Proc. of MAI. 2019. №105. С. 78-89.
  13. Nguyen D., Patel R., Zhang Y. Reinforcement Learning for UAV-5G Network Op-timisation [Electronic resource]. URL: https://arxiv.org/abs/2403.01234 (date of reference: 15.04.2025).
  14. Kozlov, D.A.; Belov, A.N. Application of the neural network methods for the forecasting of the communication channels parameters in the aviation systems // Proc. of MAI. 2020. №108. URL: www.mai.ru/science/trudy (date of address 04.03.2025).
  15. Smith A., Johnson B., Williams C. Cyclostationarity Analysis in 5G UAV Networks // IEEE Access. 2023. Vol. 11. P. 12345-12360. DOI: 10.1109/ACCESS.2023.3245678.
  16. Lee B. Deep Learning for UAV Signal Analysis: Theory and Applications. Berlin: Springer, 2022. 312 p. ISBN 978-3-031-12345-6.
  17. GOST R 57700.37-2021 Computer models and modelling. Machine learning methods. Part 37: General requirements’. Moscow: Standardinform, 2021. 24 с.
  18. ISO/IEC 23053:2021 Framework for Artificial Intelligence (AI) System Using Ma-chine Learning (ML). Geneva: ISO, 2021. 42 p.
  19. Kohavi R. A Study of Cross-Validation and Bootstrap for Accuracy Estimation and Model Selection // Proceedings of the 14th International Joint Conference on Artifi-cial Intelligence (IJCAI). 1995. Vol. 2. P. 1137-1143.
  20. ITU-T Y.3172. Framework for evaluating machine learning in next-generation net-works. ITU, 2020.
  21. Sokolov, N.I.; Mikhailov, A.V. Efficiency estimation of the hybrid signal processing methods in the aviation systems // Proc. of MAI. 2023. №118. URL: www.mai.ru/science/trudy (date of address 04.03.2025).
  22. Yakovlev, P.D.; Zimin, K.L. Application of machine learning for resource optimisation in the UAV communication network // Proc. of MAI. 2022. №116. URL: www.mai.ru/science/trudy (date of address 04.03.2025).
  23. Wang L., Li J. SCD Analysis for 5G Non-Stationary Channels // IEEE Access. 2022. Vol. 10. P. 98765-98780. DOI: 10.1109/ACCESS.2022.987654.
  24. 3GPP TR 38.901. Study on Channel Model for Frequencies from 0.5 to 100 GHz (Release 16). 3GPP, 2022. V16.1.0. 98 p.
  25. UAV Network Optimisation Dataset [Electronic resource] / Ziya M. // Kaggle. 2023. URL: https://www.kaggle.com/datasets/ziya07/uav-network-optimization-dataset (date of address: 15.04.2025).

Download

mai.ru — informational site MAI

Copyright © 2000-2026 by MAI

 Вход