The International Conference on Security, Fault Tolerance, Intelligence

Method for malicious network traffic categorisation

Artem Dremov — Fri, 10 May 2024 00:00:00 +0300

This paper aims to provide a solution for malicious network traffic detection and categorisation. In this paper we propose a semi-supervised GAN to train a discriminator model to categorise malicious traffic, as well as identify malicious and non-malicious traffic. The main goal is to achieve accurate categorisation of malicious traffic with few labelled examples.

CLASSIFICATION MODEL OF MILITARY AVIATION BASED ON NEURAL NETWORK ENSEMBLE

Volodymyr Oliinyk, Pavlo Ponochovnyy — Fri, 10 May 2024 00:00:00 +0300

The paper presents a model for classifying military aviation based on images using an ensemble of neural networks. Popular actual convolutional neural networks (CNN) served as the base models and were fined-tuned on relevant data. An ensemble of models was utilized to improve performance achieving high accuracy with reasonable compute requirements.

USAGE DIAGNOSTICS TO IMPROVE THE FAULT TOLERANCE OF THE IT STRUCTURE ELEMENT BASE

Olena, Артем — Fri, 10 May 2024 00:00:00 +0300

The article considers the idea of obtaining information about the technical conditions of the element base with the decomposition of the IT structure. Preliminary diagnosis of the components of the element base organizes neural network technologies to increase the fault tolerance of IT structures. Classification and sorting of components according to their technical condition was carried out in the MATLAB environment.

Key words: fault tolerance, neural network technologies, diagnosis, element base of the IT structure.
Fig.: 3. Tabl.:1. Bibl.: 13.

ENHANCING SYSTEM FAULT TOLERANCE THROUGH THE USE OF BLOCKCHAIN TECHNOLOGY

Роман Серебряков, Artem Volokyta — Fri, 10 May 2024 00:00:00 +0300

The paper deals with the issues of the profitability of using blockchain technology to create more enhanced fault-tolerant systems. The advantages and disadvantages of using public and private blockchain are considered, as well as the impact of this technology on distributed and centralized systems. Examples of the use of blockchain technology to increase system productivity in various spheres of human activity are given.

IMPROVING THE EFFICIENCY OF FUNCTIONS COMPUTATION IN ON-LINE MODE ON FPGA

Illya Verbovskyi, Valerii Zhabin — Fri, 10 May 2024 00:00:00 +0300

The paper considers the possibility of reducing resources and increasing the speed of computing functions in on-line mode on the basis of recursive-digital low-pass filter.

Approach to design and development of services in information systems of telecommunications providers

Fri, 10 May 2024 00:00:00 +0300

The problem of design and development of services in the information systems of telecommunications providers is considered, taking into account evaluations of the effectiveness of services, trends in the development of the IT industry and analysis of user needs from various sources. The problem is considered at the levels of service design, the development of the provider's service catalog, the provider's service package for the consumer company, and the development of the provider's services. When solving problems of design and development of services at each level, in addition to traditional parameters, new evaluations of service efficiency, development trends of the IT industry and analysis of user needs from various sources are used, including the use of consumer feedback on social media.

FORECASTING-AT-SCALE ALGORITHMS FOR PREDICTION CLUSTER WORKLOAD

Vitalii Omelchenko, Oleksandr Rolik — Fri, 10 May 2024 00:00:00 +0300

The paper deals with the issue of predicting workloads in a cluster to use in proactive scaling of computing resources. Forecasting-at-scale algorithms Prophet and Greykite for forecasting time series are considered, their accuracy and universality are evaluated.