Study of the combined variant of determination of attacks using neural network technologies
DOI:
https://doi.org/10.34185/1562-9945-3-140-2022-08Keywords:
attack, category, class, perceptron, neuro-fuzzy network, self-organizing Kohonen map, combined variant, quality parametersAbstract
The modern world is impossible to imagine without computer networks: both local and global; therefore, the issue of network security is becoming increasingly topical. Currently, methods of detecting attacks can be strengthened by using neural networks, which confirms the relevance of the topic. The aim of the study is a comparative analysis of the quality parameters of network attacks using a combined variant consisting of different neural networks. As research methods used: neural network; multilayer perceptron; Kohonen's self-organizing map. The software implementation of the Kohonen self-organizing map is carried out in Python with a wide range of modern standard tools, creation of a multilayer perceptron and a fuzzy network - using Neural Network Toolbox packages, and Fuzzy Logic Toolbox system MatLAB. On the created neural networks separately and on their combined variant researches of parameters of quality of definition of network attacks are carried out. It was determined that the error of the first kind was 11%, 4%, 10% and 0%, the error of the second kind - 7%, 6%, 9% and 6% on the fuzzy network, multilayer perceptron, self-organizing Kohonen map and their combined version, respectively, which proves the feasibility of using the combined option.
References
Branitskiy A.A. Obnaruzhenie anomalnykh setevykh soedineniy na osnove gibridizatsii metodov vychislitelnogo intellekta (Extended abstract of PhD disserta-tion). St. Petersburg, Russia, 2018.
Emelyanova Yu. G., Talalaev A. A, Tishchenko I. P, Fralenko V. P. Neural network technology for detecting network attacks on information resources. Software sys-tems: theory and applications, 2011. № 3(7). С. 3-15.
Mustafaev A. G. Neural network system for detecting computer attacks based on network traffic analysis. Security questions. Вопросы безопасности, 2016. № 2. С. 1-7. DOI: 10.7256.2409-7543.2016.2.18834
Pakhomova V. M., Konnov M. S. Research of two approaches to detect network attacks using neural network technologies. Science and Transport Progress, 2020, 3(87), 81-93. URL: https://doi.org/10.15802/stp2020/208233
Network attack detection technologies. Brest State Technical University. URL: https://www.bstu.by/~opo/templates_c/%25%25A1%5EA14%5EA14FF5EA%25%25index.html.php
Frolov P. V., Chukhraev I. V., Grishanov K. M. Application of artificial neural net-works in intrusion detection systems. System administrator, 2018. 9(190). Retrieved from samag.ru/archve/article/3724
Amini, M., Rezaeenour, J., Hadavandi, E. A Neural Network Ensemble Classifier for Effective Intrusion Detection Using Fuzzy Clustering and Radial Basis Function Networks. International Journal on Artificial Intelligence Tools, 2016. 25(02), 1-32. DOI: https://doi.org/10.1142/s0218213015500335
Dhangar K., Kulhare D., Khan A. A Proposed Intrusion Detection System. International Journal of Computer Applications. 2013. Vol. 65, N 23. р.р. 46-50.
NSL-KDD dataset. URL: https://www.unb.ca/cic/datasets/nsl.html
Pakhomova V. M., Bikovska D. G. Investigation of multilayer neural network parameters for determination of R2L category network attacks. Modern engineering and innovatite technologies. Germany, Karlsruhe: Sergeieva&Co, «ISE&E», 2021. № 18-02. рр. 39-43. DOI: 10.30890/2567-5273.2021-18-02-059
Saied A., Overill R. E., Radzik T. Detection of known and unknown DDoS at-tacks using Artificial Neural Networks. Neurocomputing, 2016. 172, 385-393. DOI: https://doi.org/10.1016/j.neucom.2015.04.101
Zhukovyts’kyy I. V., Pakhomova V. M. Identifying threats in computer net-work based on multilayer neural network. Science and Transport Progress, 2018. 2(74), 114-123. DOI: https://doi.org/10.15802/stp2018/130797
Zhukovyts’kyy I. V., Pakhomova V. M., Ostapets D. O., Tsyhanok O. I. Detec-tion of attacks on a computer network based on the use of neural network complex. Наука та прогрес транспорту. 2020. № 5(89). 68-79. URL:
