Development of an intelligent software module based on neural networks for visualization and automatic analysis of characteristics of information signal spectra and destructive impacts

   The paper proposes a method for implementing an intelligent software module based on neural networks for visualization and automatic analysis of the characteristics of information signal spectra and destructive influences. A procedure is given for selecting and creating a machine learning model that allows identifying the presence of destructive effects and its signs. The neural network model is trained in a cloud environment using a GPU. A specialized library was used to train the neural network model. Training is implemented on the parameters of the neural network, which characterize the number of times the dataset passes through the neural network in the forward and reverse directions. As a result of training, graphs were obtained characterizing the dependence of errors on the number of epochs of model training. An algorithm for identifying power characteristics, the width of the spectrum of destructive effects in the spectrum of an information signal, and an intelligent software module (script) for their automatic analysis has been developed. The mechanism for analyzing the spectrum width and power of destructive influence in the spectrum of an information signal includes software and machine learning models for recognizing text in images.

1. Khokhlov, N. S. Analysis of some information security vulnerabilities of the wireless communication system of the DMR standard / N. S. Khokhlov, S. V. Kanavin, I. V. Gilev // Bulletin of the Voronezh Institute of the Ministry of Internal Affairs of Russia. – 2022. – No. 1. – P. 9–17. – ISSN 2071-3584.

2. Khokhlov, N. S. Model of countering threats of information destruction in special-purpose communication systems under destructive influences / N. S. Khokhlov, S. V. Kanavin, I. V. Gilev // Bulletin of the Voronezh Institute of the Ministry of Internal Affairs of Russia. – 2023. – No. 1. – P. 106–117. – ISSN 2071-3584.

3. OCR on Raspberry Pi using Tesseract. – URL: https://microkontroller.ru/raspberry piprojects/opticheskoe-raspoznavanie-simvolov-v-raspberry-pi-s-pomoshhyu-tesseract (date of access: 10/18/2023).

4. Sirota, A. A. Analysis of algorithms for searching objects in images using various modifications of convolutional neural networks / A. A. Sirota, E. Yu. Mitrofanova, A. I. Milovanova // Bulletin of Voronezh State University. Series: System analysis and information technologies. – 2019. – No. 3. – P. 123–137. – ISSN 1995-5499.

5. Chorbaa, N. A. Comparative analysis of methods for detecting objects on radar images using neural networks / N. A. Chorbaa, A. Tu. Le, I. M. Tolstoy // Scientific result. Information Technology. – 2020. – Vol. 5, No. 4. – P. 15–25. – ISSN 2518-1092.

6. Application of artificial neural networks for the analysis of multispectral images / M. Yu. Alyes, E. A. Antonov, A. I. Kalugin, M. R. Zaripov // Optical Journal. – 2021. – Vol. 88, No. 8. – P. 48–53. – ISSN 1023-5086.

7. Chernigovsky, A. V. Neural networks as a tool for analyzing network traffic / A. V. Chernigovsky, M. V. Krivov // Bulletin of the Angarsk State Technical University. – 2019. – No. 13. – P. 151–157. – DOI: 10.36629/2686-777x-2019-1-13-151-157.

8. On the issue of ensuring secure access to information systems using biometric authentication based on a fuzzy image of the user’s personality and neural network transformations / O. I. Bokova, S. V. Kanavin, N. S. Khokhlov [et al.] // Bulletin of the Dagestan State Technical University. Technical science. – 2023. – No. 50 (4). – P. 75–84. – DOI: 10.21822/2073-6185-2023-50-4-75-84.

9. Fisun, V. V. Artificial intelligence for managing information security of critical information infrastructure objects: monograph / V. V. Fisun. – Moscow : Rusigns, 2023. – 360 p. – ISBN 9785436597232.

10. Klimov, S. M. Countering computer attacks. Methodological foundations. / S. M. Klimov, M. P. Sychev, A. V. Astrakhov. – Moscow : MSTU named after N. E. Bauman, 2013. – 108 p.

11. 11 Shelukhin, O. I. Network anomalies. Detection, localization, forecasting : monograph / O. I. Shelukhin. – Moscow : Hotline – Telecom, 2020. – 447 p. – ISBN 978-5-9912-0756-0.

12. Albon, K. Machine learning using Python. Collection of recipes : Transl. from English / K. Elbon. – Saint Petersburg : BHV-Petersburg, 2019. – 384 p. – ISBN 978-5-9775-4056-8.

13. FSH8 portable spectrum analyzer. – URL: https://www.samarapribor.ru/main/fsh8.html (date of access: 10/18/2023).

14. BL-YOLOv8 : An Improved Road Defect Detection Model Based on YOLOv8. – URL: https://www.mdpi.com/1424-8220/23/20/8361#B19-sensors-23-08361 (date of access: 10/18/2023).

15. Presnetsov, A. M. Development of a hardware and software complex for monitoring production activities using the YOLOV8 neural network / A. M. Presnetsov, A. P. Tyurin // Intelligent systems in production. – 2023. – Vol. 21, No. 2. – P. 140–151. – ISSN 1813-7911.

16. Certificate of state registration of the computer program RU, No. 2023688381 dated December 21, 2023. Intelligent software module for automated analysis of the characteristics of information signal spectra and destructive influences. / R. V. Rtishchev, I. V. Gilev, S. V. Kanavin, N. S. Khokhlov // No. 2023687782 ; application 12. 08. 2023, Bulletin. No. 1. – 1 p.

17. Postolit, A. V. Fundamentals of artificial intelligence in examples in Python. Self-instruction manual / A. V. Postolit. – 2^nd ed., revised. and additional. – Saint Petersburg : BHV-Petersburg, 2024. – 448 p. – ISBN 978-5-9775-1818-5.