Oleg BARABASH, DSc (Engin.), Prof.
ORCID ID: 0000-0003-1715-0761
e-mail: bar64@ukr.net
National Technical University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”, Kyiv, Ukraine
Andrii MAKARCHUK, PhD Student
ORCID ID: 0000-0002-6422-7488
e-mail: andreymakarh2@gmail.com
National Technical University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”, Kyiv, Ukraine
DOI: https://doi.org/10.17721/AIT.2024.1.06
Abstract
B a c k g r o u n d . Functional The functional stability of distributed systems is becoming increasingly significant with the advancement of information technologies. Consequently, the formalization of this concept has gained relevance. Mathematical formalization of functional stability in the form of its indicators and criteria has been underway for several decades. Functional stability indicators play a key role in this process, and many such indicators have already been formulated. However, a major drawback of most of these indicators is that they are not only computationally complex but also dependent on numerous other parameters, or they fail to comprehensively describe the functional stability of the distributed system in question. In this paper, a new functional stability indicator is introduced that avoids the second of these drawbacks. The first drawback is addressed through the use of an estimation method based on multivariate polynomial regression.
M e t h o d s . The study utilized methods of computer modeling and approximation techniques.
R e s u l t s . A modification of an existing indicator, known in the literature as the probability of reliability, was chosen as the method for developing the new functional stability indicator. By making certain assumptions and applying transformations, a measure was obtained that possesses certain desirable properties, namely: this measure lies strictly within the interval from zero to one, and the larger it is, the more functionally stable the distributed system under consideration can be deemed. However, the resulting functional stability indicator requires extensive calculations, prompting an attempt to estimate this indicator using approximation methods. This study explored the potential of applying multivariate polynomial regression. According to computer modeling, to achieve an average accuracy of two percent, it is sufficient to use a five-dimensional polynomial regression of the fourth degree. Increasing the degree of the five-dimensional regression model beyond this does not result in significant error reduction.
C o n c l u s i o n s . The functional stability indicator introduced in this study provides a convenient means for investigating the functional stability of distributed systems. However, it demands a significant amount of computation. For this reason, a method for estimating the introduced functional stability indicator has been presented, which allows for relatively accurate computation of this indicator.
K e y w o r d s : functional stability, indicator, approximation, optimization, regression, multivariable functions.
Published
2024-12-20
How to Cite
Oleg BARABASH, Andrii MAKARCHUK “ DEVELOPMENT OF A NEW INDICATOR OF FUNCTIONAL RELIABILITY AND ITS EVALUATION USING MULTIVARIABLE POLYNOMIAL REGRESSION” Advanced Information Technology, vol.1(3), pp. 59–66, 2024
Issue
Advanced Information Technology № 1 (3), 2024
Section
Applied information systems and technology
References
Abdullayev, F. G., Ozkartepe, P., Savchuk, V. V., & Shidlich, A. L. (2019). Exact constants in direct and inverse approximation theorems for functions of several variables in the spaces Sp. Filomat, 33(5), 1471–1484. https://doi.org/10.2298/FIL1905471A
Antonevych, M., Didyk, A., & Snytyuk, V. (2019). Optimization of functions of two variables by deformed stars method. In O. Yudin (Ed.). 2019 IEEE International Conference on Advanced Trends in Information Theory (ATIT) (pp. 475–480). Taras Shevchenko national university of Kyiv https://doi.org/10.1109/ATIT49449.2019.9030475
Barabash, O. V. (2004). Construction of functionally stable distributed information systems. National Academy of Defense of Ukraine [іn Ukrainian]. Barabash, O. V., & Kravchenko, Y. V. (2002). Functional stability – a property of complex technical systems. Collection of Scientific Papers of the National Academy of Defense of Ukraine, 40, 225–229 [In Ukrainian].
Barabash, O. V., Obidin, D. M., Salanda, I. P., & Makarchuk, A. V. (2024). Comparative analysis of the approximation of a probabilistic functional stability indicator using Bernstein polynomials and feedforward neural networks. Sviaz, 3, 32–37 [іn Ukrainian]. https://doi.org/10.31891/2219-9365-2024-77-6
Cormen, T. H., Leiserson, C. E., Rivest, R. L., & Stein, C. (2009). Introduction to algorithms (3rd ed.). The MIT Press.
Kravchenko, Y., Leschenko, O., & Mykus, S. (2016). Functional stability of information and telecommunication systems. East European Scientific Journal,2(6), 47–52. https://doi.org/10.2298/FIL1905471A
Kovalchuk, D., Kravchenko, Y., Starkova, O., Herasymenko, K., Tarasenko, N., & Riabtsev, V. (2020). Development of recommendations for the implementation of virtualization concepts in modern networks. In D. Ageyev (Ed.). 2020 IEEE International Conference on Problems of Infocommunications, Science and Technology (PIC S&T) (pp. 797–802). National Aerospace university “KhAI. https://doi.org/10.1109/PICST51311.2020. 9467918
Makarchuk, A., Kal’chuk, I., Kharkevych, Y., & Kharkevych, G. (2022). Application of trigonometric interpolation polynomials to signal processing. In O. Yudin (Ed.). 2022 IEEE 4th International Conference on Advanced Trends in Information Theory (ATIT) (pp. 156–159). Taras Shevchenko national university of Kyiv. https://doi.org/10.1109/ATIT58178.2022.10024182
Mashkov, O. A., Murasov, R. K., Kravchenko, Y. V., Dakhno, N. B., Leschenko, O. A., & Trush, O. V. (2021). Optimal forecast algorithm based on compatible linear filtration and extrapolation. Mathematical Modeling and Computing, 8(2), 157–167. https://doi.org/10.23939/mmc2021.02.157
Миронюк, М. Ю., Майстров, О. О., Мусієнко, А. П., & Макарчук, А. В. (2024). Аналіз побудови інтелектуальної інформаційної системи на основі поняття функціональної стійкості. Зв’язок, 1, 3–8. https://doi.org/10.31673/2412-9070.2024.010308
Shidlich, A. L. (2013). Approximations of certain classes of functions of several variables by greedy approximants in the integral metrics. Institute of Mathematics of NAS of Ukraine.
Snytyuk, V., Antonevich, M., & Didyk, A. (2020). Optimization of fire monitoring system using the deformed stars method. Information Systems and Technologies Security, 1(2), 60–66. https://doi.org/10.17721/ISTS.2020.1.60-66