Numerical Laplace transform and inverse Laplace transform is a challenging task in queueing theory and others probability models. A double transformation approach is used to find stable, accurate, and computationally efficient methods for performing the numerical Laplace and inverse Laplace transform. To validate and improve the inversion solution obtained, direct Laplace transforms are taken of the numerically inverted transforms to compare with the original function. Algorithms provide increasing accuracy as precision level increases. The most promising methods were applied to computational probability models, when there are no closed-form solutions of the Laplace transform inversion. The computational efficiency compared to precision levels is demonstrated for different service models in M/G/1 queuing systems.

Aim. To analyse the factors that may affect the choice of the method(s) for assessing the failure rate of system components when using the results for reliability analysis and probabilistic safety analysis. Methods. The paper uses methods of mathematical analysis, probability theory. Results. The paper presents the results of applying a number of methods to a sample of N = 23 (groups) and samples of N = 12 and N = 6. Samples N = 12 and N = 6 are made up of a static series of the initial sample (N = 23) according to the following rule adopted for the purpose of demonstrating the method: for N = 12, all even, for N = 6, every 4th, that' ensured the “similarity” of the samples. Two variants, i.e. z = 0, t = 11.2 and z = 4, t = 11.2, were considered as specific data. Conclusions. The paper presents a brief overview of the practical methods for assessing the failure rate characteristics of a nuclear power unit’s system components. The authors highlight the specific features of the presented approaches, provide recommendations for their application in the preparation of input data for analysis. For the case of no failures in a group, it is recommended using the Jeffreys noninformative method. Relibaility analysis should use the method of moments, the V.B. Morozov’s method and/or the bootstrap method.

Aim. Evaluating distribution parameters based on small samples is an unconventional problem in itself. Solving it by maximizing the likelihood function may produce highly biased results. The paper analyses the properties of some small-sample estimates of beta distribution parameters of the 1-st kind. Methods. The comparison of small-sample estimates of beta distribution parameters using various methods involved simulation with the number of tests N = 10⁴. Results. Parameter estimation using the maximum likelihood method does produce a highly biased result for small samples. The bootstrap method, as compared to the maximum likelihood method, produces less biased estimates with a smaller variance. The most acceptable (close to the initial values) result was obtained using the mathematical expectation (or median) and variance. Conclusion. For small samples, no particular method of parametre estimation can be recommended. The neural network analysis appears to be the best suited for small samples. Neural network integration of a number of methods of estimation may significantly improve its accuracy.

Despite the widespread use of information and information modeling, the reliability of information has not yet been thoroughly studied. Aim. The paper suggests a method for analysing the reliability of information as a new type of reliability. A model of information reliability parameters has been defined. Methods. The paper uses methods of categorical qualitative and comparative analysis. Results. The research has defined special features of the reliability of information that distinguish it from other types of reliability. Conclusion. The proposed method of analysing the reliability of information allows improving the quality and reliability of information processes and the reliability of information processing results.

Aim. To examine the assessment of cyber security of intelligent transportation systems with a multi-level information protection system. For the purpose of preventing attacks aimed at capturing information in a multi-level protection system, it is proposed to implement a fully connected core of the local area network (at all levels). This enables random trusted paths that are dismantled upon transmitting the information, i.e. are limited in their time of existence. Communicating information along randomly selected trusted limited-lifetime paths complicates internal attacks that aim to capture network entities involved in the communication of critical information. Methods. The paper uses methods of mathematical analysis, graph theory, and probability theory. Results. The authors examine a model of traffic capture by an attacker, whereas the defender uses random pathing in a fully connected network. A model of information protection in a multi-level information protection system was assessed. It is proposed using dynamically organised trusted paths in fully connected environments when designing a multi-level protection system to prevent interception of information flows. Conclusion. The proposed technique allows estimating the decrease in the probability of vertex capture in a fully connected network, as well assessing the probability of vertex capture depending on the duration of message transmission.

Aim. To use the case of one of the ELECTRE methods to identify and demonstrate the applicability of the development of indexes for pairwise comparison of alternatives (RIPSA) for selecting automated information systems (AIS) taking into account criteria associated with AIS cybersecurity and cyber risk insurance. Methods. The paper uses the ELECTRE I method that is well-described in well-known publications. The ELECTRE group of methods belongs to the RIPSA approach that aims to compare multi-criteria alternatives. One of the Russian pioneers of this approach was the RAS member Larichev O.I. The introduced criterion called the “three-digit majority criterion of AIS cybersecurity taking into account insurance” uses elements of three-digit majority logic. Results. The paper shows the applicability of the RIPSA approach, ELECTRE I specifically, for selecting AIS as regards the examined criteria associated with AIS cybersecurity, taking into account cyber risk insurance. A modified criterion has been developed that is new as regards cyber security in the context of cyber liability insurance and is based on a three-digit majority logic. Specifically, it allows expressing and taking into account in a logical and mathematical form some features of AIS insurance protection given the inherent cyber risks, as well as the ability of AIS to independently withstand the three categories of cyber attacks. Conclusion. The three-digit majority logic-based technical and economic quality criterion is applicable to not only technological and organisational parameters or characteristics of AIS cyber security, but the financial and economic parameters or security characteristics of insured AIS as well. The ELECTRE methods of developing pairwise alternative comparison indexes are applicable both in the analysis, organisational and technological measures to reduce various AIS-related cyber risks, and in associated insurance protection. The paper shows that the developed integral criterion, namely the three-digit majority logic for multi-criteria alternatives, allows using the ELECTRE methods with two criteria less.

The paper presents a system for detecting faults in distributed software and hardware systems that is based on a set of intelligent technologies. The method combines dynamic testing (fuzzing) enhanced with large language models, as well as analysis of vulnerability patterns of the well-known MITRE and OWASP knowledge bases to identify software errors that enable potential attacks. The proposed architecture promptly diagnoses failures and faults, localizes their causes and automatically escalates the incident to the system administrator. The practical significance of the solution is confirmed experimentally in terms of such parameters as the average error detection time, code coverage, and the number of detected defects.

Aim. To formalise the scientific problem of quantifying the operational stability of critical information infrastructure facilities exposed to information security threats. Methods. Cognitive methods: systems analysis, induction and deduction, analysis of a scientific problem, formalisation of scientific knowledge, construction of hypotheses. Operational methods: abstraction, specification, comparison, generalisation, analogy, simulation, expert evaluation. Results. The paper substantiates the relevance, defines – both verbally and formally – the scientific problem of quantifying the operational stability of critical information infrastructure facilities, suggests indicators for assessing the input data and the result. Conclusions. The authors systematically analysed the problem, which allowed substantiating the relevance of a potential solution, formalising it, substantiating the choice of the controllable and uncontrollable factors for assessing stability, defining the restrictions, suggesting the method for taking into account the dynamics of critical information infrastructure exposed to information security threats.

Objective. Analysis of the evolution and role of the JSC “NIIAS” Training Center in the context of the digital transformation of the railway industry, overcoming the “competence vacuum” and forming personnel potential for ensuring technological sovereignty. Definition of methodological foundations and promising directions of the Center’s activity as an integrator of advanced competencies. Methods. The study uses historical and logical analysis of the Training Center’s development stages, a systematic approach to considering its role in the industry ecosystem “RZD – NIIAS – Universities”, as well as the methodology of problem- oriented learning based on the principle of “teaching to understand the ocean rather than giving fish”. Methods of cascade knowledge transfer and modeling of integration processes in the educational environment are used. Results. The key stages of the transformation of the Training Center’s mission have been identified and characterized: from eliminating systemic knowledge gaps of personnel to proactively forming a unified field of competencies and creating professions of the future. The unique methodology of “live” learning is revealed, which minimizes the gap between educational programs and the real requirements for the implementation of innovations, such as “virtual coupling” and robotic complexes. A cascade integration model has been developed and presented, ensuring knowledge synchronization between technology developers (JSC “NIIAS”), the customer (JSC “RZD”) and educational institutions. New professional retraining programs in robotization, scheduled for launch in 2026, are announced: “Business Analyst for Robotization of Processes in Railway Transport” and “Auditor for Robotization of Production Processes in Transport”. Conclusion. The activity of the JSC “NIIAS” Training Center has evolved towards creating a system of advanced personnel training, which allows not only to respond to current challenges but also to proactively shape the future of railway transport. The Center is positioned as a key element of the intellectual protection of critical infrastructure, whose methodology and integration model contribute to overcoming the “competence vacuum”, improving traffic safety and ensuring the technological sovereignty of the industry. The creation of a unified digital educational and production ecosystem (“Digital Contour of RZD Personnel Training”) is recognized as a promising direction for development.