The International Conference on Security, Fault Tolerance, Intelligence

USING LLMs IN FINANCIAL BUDGETING APPLICATIONS

Daniil Soloviov — 2024-09-28

The article considers the potential of using large language models (LLMs) in financial budgeting systems. It will be examined how language models can be used to automate tasks, improve accuracy, improve efficiency, and increase transparency in financial budgeting. We will note some of the problems and limitations associated with the use of LLMs in financial budgeting and how these problems can be solved by applying and configuring a large language model for analysis, planning and cost control tasks in the financial field.

COMPUTER VISION BASED AUTHENTICATION MODEL WITH SPOOFING PROTECTION

Serhii Hatsan — 2024-12-04

The paper offers an example of the implementation of a computer vision subsystem for user identification with protection against spoofing. The basic models for this subsystem were the Siamese neural network based on SqueezeNet for face recognition and the network based on ResNet for face anti-spoofing.

A METHOD OF PLAYING GO USING ARTIFICIAL INTELLIGENCE

Oleksandr Morozov-Leonov — 2024-09-28

The paper deals with the issue of creating a way to play Go using artificial intelligence. The method that combines modern principles of artificial intelligence is considered, taking into account certain limitations caused by the target hardware and software.

DEVELOPMENT OF MLOPS PIPELINE TO SUPPORT AI SYSTEMS

Volodymyr Rusinov — 2024-09-28

Relevance of the research topic. Scientific and engineering literature shows that the use of Machine Learning Operations and Machine Learning Pipeline (MLOps) methods in the creation of AI-platforms for embedded systems allows to improve the efficiency of the development and implementation of artificial intelligence. MLOps provides the ability to automate the processes of developing, training and operating machine learning models, providing a quick response to changes in the environment or input data. This helps to reduce the time to deploy and maintain the system as a whole.

To measure the overall usability of the system with a certain degree of accuracy we will introduce several metrics that will help us establish the parameters of the system and help us compare it with the alternative solutions.

Target setting. Deployment of AI software must be accompanied by a thorough analysis of the target system. In this article a suite of tests will be performed to quantify the performance of the system. To outline the aspects of the system that need to be modified, several DORA metrics will be compared.

Actual scientific researches and issues analysis. Current scientific research suggests that MLOps is a novel way of deployment of AI systems that inherits its methods from DevOps, that has shown positive results in various software systems. Various articles propose different variation of tools and CI/CD pipeline designs that enable AI tasks to be performed at scale and increase the overall fault tolerance of the system. These solutions also propose a method to automate any processes that would require additional human resources to achieve, such as monitoring, logging and reconfiguration.

The research objective. The goal of the article is to investigate the MLOps principles and quantify their impact on the deployment.

Presentation of the main material. Machine Learning Operations (MLOps) originates as a concept based on DevOps methods. It is an attempt to create a collection of procedures that combine tools, deployment procedures, and workflows to quickly and affordably build machine learning models [1]. MLOps recommends automating and controlling every step of creating and implementing machine learning systems. [2] MLOps practices are perceived as a complex coordination of a set of software components that are used in an integrated manner to perform at least five tasks: data collection; data conversion; continuous learning of the machine learning model; continuous implementation of the model; and presentation of results to the end user. This is achieved by using MLOps principles.

Conclusion. Machine Learning Operations is a DevOps concept that is proven to create a streamlined process for creating machine learning models. It involves automating and monitoring each step of the process, including data collection, transformation, training, implementation, and end-user results presentation. To ensure the effectiveness of MLOps, organizations should choose appropriate tools and technologies, such as cloud platforms like AWS or Microsoft Azure.

Continuous integration and Continuous Delivery are a crucial for efficient deployment and testing for application readiness. It automates the process of deployment through creation of a pipeline that continuously tests and builds ML software systems. Machine Learning (ML) systems require an experimental model, which is analyzed, cleaned, and pre-processed before training, testing, and verification. The deployment phase ensures model integration and stability, with continuous monitoring by MLOps engineers.

Testing is performed using DORA metrics. A comparison of the solution proposed in this article with other solutions shows. The results table compares metrics for a modified approach, focusing on GitHub upload and lead time to changes, comparing them for further testing. The proposed method is less sophisticated than its alternatives, however it is more lightweight therefore requiring less time to build and reconfigure after faults.

Uninvestigated parts. To increase the efficiency of this approach, additional research may be required in the fields of container optimization, AI optimization as well as networking and administration. This article aims to explore the ways MLOps is used and to propose a solution that is lightweight and retains the most important properties and adheres to the outlined principles. However, it is important to note that in real-life scenarios, additional changes may be required to increase the fault tolerance of the system and to increase its capabilities to scale.

USING DEEP LEARNING FOR VIDEO CONTENT GENERATION

Sofiia Sytnik — 2024-09-28

The paper discusses the use of deep learning methods for automated video content generation. The advantages and disadvantages of various neural network architectures such as Convolutional Neural Networks (CNNs), Generative Adversarial Networks (GANs), and transformers are analyzed. Special attention is paid to the use of these technologies in various fields, including cinematography, the gaming industry, and video surveillance.

Keywords: deep learning, video generation, CNN, GAN, transformers

Relevance of the research topic. Deep learning, forming the backbone of many modern artificial intelligence applications, is critically acclaimed for its efficacy in handling complex and large datasets, particularly in the field of video generation. Its capabilities extend to enhancing the realism and personalization of generated content, making it essential to explore these technologies further for industrial applications.

Target setting. The aim is to evaluate the potential and efficacy of deep learning frameworks in automating video content generation across various applications.

Actual scientific researches and issues analysis. Recent studies have underscored the effectiveness of neural networks, particularly GANs and transformers, in producing high-quality, realistic video sequences that are hard to distinguish from real footage. However, challenges like high computational costs and the need for large datasets pose significant hurdles.

Uninvestigated parts of general matters defining. This article is devoted to studying the integration and optimization of deep learning technologies in high-demand video generation tasks, especially in environments requiring dynamic content creation.

The research objective. To analyze the suitability of deep learning technologies in enhancing video generation processes. Consider the benefits and drawbacks of these technologies in both centralized and distributed systems.

The statement of basic materials. Deep learning technology, underpinning sophisticated video generation frameworks, offers substantial promise for constructing highly immersive and interactive media [1]. Features like automated scene generation, dynamic object integration, and real-time rendering capabilities allow for significant enhancements in video quality and interaction.

Deep learning for video generation. The capacity of deep learning systems to simulate realistic animations and interactions in video content is crucial for sectors like virtual reality (VR), video games, and online education [2]. Techniques like neural style transfer, motion capture, and facial recognition are employed to produce videos that are not only visually appealing but also contextually appropriate.

Public and private frameworks. The adaptability of deep learning models in public and private settings also varies significantly. Publicly available models can be fine-tuned for generic tasks, while private, bespoke models are tailored for specific enterprise needs, balancing cost-efficiency with computational demand.

Blockchain in video generation. While not the focus of this paper, integrating blockchain for enhancing security and copyright management in the distribution of generated video content is a viable consideration.

Other approaches to enhance video generation. In addition to deep learning, other computational techniques like cloud computing and edge computing are being explored to distribute the processing load and reduce latency in video generation tasks.

Application areas. Areas where deep learning significantly impacts include:

1.Cinematography: Automating script-to-screen processes, enhancing visual effects with fewer manual interventions.

2.Gaming: Generating dynamic game environments that react to player actions in real-time.

3.Surveillance: Improving the accuracy and reliability of surveillance systems through enhanced object detection and scenario simulation.

Generative Adversarial Networks (GANs)[image 1] are increasingly being utilized in various application areas due to their ability to generate high-quality, realistic images and videos. In video production, GANs are particularly valuable for creating lifelike animations and effects seamlessly integrated into live-action footage. This application is crucial in fields such as film and television production, where the demand for high-quality visual content is constantly rising.

In the domain of medical imaging, sparse autoencoders facilitate the enhancement of image clarity and detail, aiding in more accurate diagnosis and analysis. The ability to extract significant features from medical scans while ignoring irrelevant data reduces computational overhead and improves processing times.

The regularization term in the loss function helps control the sparsity level, ensuring that the network does not overfit and that it generalizes well to new, unseen data. This aspect is crucial in applications like facial recognition and anomaly detection in surveillance systems, where distinguishing between normal and unusual patterns accurately can be vital.

These applications demonstrate the versatility and potential of both GANs and sparse autoencoders across various high-impact fields, leveraging their unique capabilities to improve the efficiency and quality of outcomes in industry-specific challenges.

Implementation problems. Implementing deep learning technologies for video generation comes with several notable challenges. These issues must be addressed to harness the full potential of AI in enhancing video content creation effectively. The section below outlines the primary implementation challenges that developers, organizations, and researchers face when integrating deep learning models into video generation workflows.

Deep learning models, particularly those used for generating and processing video content, require substantial computational power. The training and operational phases of these models often need the use of GPUs or specialized hardware like TPUs, which can handle massive parallel processing tasks necessary for handling video data. The cost of such hardware is not trivial and represents a significant investment for startups and even some larger enterprises. Additionally, the energy consumption associated with running these powerful machines continuously is considerable, impacting operational costs and environmental footprint.

Video content often contains sensitive information. When implementing deep learning for video generation, particularly in areas like surveillance, healthcare, or personalized media, ensuring the privacy and security of the data is paramount. Compliance with international data protection regulations (e.g., GDPR or HIPAA) is necessary to protect individual privacy rights and prevent data breaches. Techniques like data anonymization, secure data storage, and encrypted data transmission become crucial in such implementations.

Deep learning models must scale efficiently to accommodate the vast amounts of video data generated daily and handle peak load times without performance degradation. Additionally, these models should be flexible enough to integrate seamlessly with existing digital asset management systems, requiring compatibility with various software and hardware configurations. Scalability not only pertains to data handling but also to the ability to maintain performance as the network architecture scales up or as the number of users increases.

Despite the advancements in AI, creating algorithms that can effectively handle the complexity and diversity of real-world video scenes remains challenging. Algorithms must be robust enough to deal with variations in video quality, lighting conditions, and unexpected environmental elements. They also need to be adaptable to new and evolving content without requiring extensive retraining or manual adjustments. This adaptability is crucial for applications that rely on real-time video analysis, such as autonomous driving and real-time surveillance.

Many industries equipped with older video management systems face significant challenges when integrating modern deep learning solutions. These legacy systems often are not designed to handle the high throughput and dynamic data processing demands of AI-based tools. Upgrading these systems to be compatible with new technologies involves not only technical changes but also organizational and workflow adjustments, which can be costly and time-consuming.

The complexity of deep learning models necessitates a high level of expertise in both software development and machine learning theory. There is a continuous need for skilled personnel who can develop, maintain, and upgrade these systems. The shortage of qualified experts can be a barrier to adoption, especially in regions or sectors where the tech industry is not as developed.

Conclusions. Deep learning technologies hold the potential to redefine the norms of video content generation across various sectors. By addressing the existing challenges and harnessing the capabilities of AI, the landscape of media and entertainment, as well as surveillance and education, will continue to evolve, making content more engaging, personalized, and accessible.

This expanded narrative provides a thorough exploration of how deep learning technologies are being integrated into video generation, highlighting the profound impact these tools have across various industries. By addressing both the technological advancements and the associated challenges, the article offers a comprehensive view of this rapidly evolving field.

Kostin,Pavlov-ENTELLIGENT SYSTEM FOR PREDICTING CRYPTOCURRENCY PRICES BASED ON STATIC DATA

Kostin Denis — 2024-09-28

The article discusses the principle of predicting cryptocurrency prices using a system based on machine learning algorithms and static data. The advantages and disadvantages of using machine learning algorithms for dynamic prediction of time series and prices are examined.

PSYCHOLOGICAL SUPPORT ASSISTANT BASED ON FINE-TUNED LLAMA 3 MODEL

Vitalina Aslanova — 2024-12-04

The paper presents an intelligent dialogue assistant for psychological support. The system is based on modern natural language processing and machine learning technologies, using libraries from Hugging Face to work with language models. Algorithms have been implemented to optimize fine-tuning.

INCREASING THE EFFICIENCY OF AI TASK DEPLOYMENT FOR CLOUD-EDGE ENVIRONMENTS

Volodymyr Rusinov — 2024-09-28

Relevance of the research topic. Contemporary literature on AI task deployment states the importance of MLOps principles and the ways of implementing them without optimizing the individual steps. Cloud-Edge hybrid architecture requires specific measures to be taken in order for these principles to be applicable to IoT devices, which have low resources. Introducing the same software to such systems may cause increased faults and decrease uptime. To modify the pipeline, a suite of techniques is introduced. Measurements such as time and disk size utilization are made to quantify the impact on the system.

Target setting. Different articles show variety in CI/CD process implementation and require analysis and modification of its steps in order to increase the efficiency and usability on different platforms. To fully utilize the capabilities of IoT systems, modifications are required that will make the images more lightweight and the pipeline faster.

Actual scientific researches and issues analysis. Current scientific research shows that CI/CD is widely used and researched in order to be applied to AI problems. However, there is much less literature on how to optimize the CI/CD pipelines and its components. These articles propose different optimization techniques that apply to different areas of research show promise and can be usable in AI systems.

Uninvestigated parts of general matters defining. Contemporary solutions provide minimal insight into the performance of such systems in comparison with other approaches. MLOps principles have not been put into testing in end-to-end pipelines to quantify their performance and outline areas of improvement. General improvement of each step is required to launch the pipeline efficiently on IoT/Edge hardware and to fully automate it using Cloud.

The research objective. The goal of the article is to investigate methods of improving the deployment task of MLOps pipelines by increasing the uptime and decreasing build times.

Presentation of the main material. The foundation of this approach is to create an efficient CI/CD pipeline. Continuous integration (CI) is a process that integrates code changes into the existing code base to resolve conflicts between changes made by different developers. It includes build and test phases that involve automatic code generation and testing. The planning and coding steps are tied to CI, making deployment more efficient.
Container optimization in this context includes a number of measures aimed at reducing the size and resource usage of the container. The purpose of optimization is to accelerate the construction and deployment speed of container applications at the network edge, as well as to reduce the use of system resources.

Conclusion. In this paper, we proposed a methodogical approach to improvement of AI-systems. MLOps practices ensure the effectiveness and quality of artificial intelligence solutions over time. The core of this approach is creating an effective CI/CD pipeline, which includes continuous integration (CI), planning and coding phases, and Continuous Delivery (CD), which keeps applications ready for production deployment.

Container optimization is a method that considers individual system elements and their impact on its functioning. It aims to speed up the build and deployment of containerized applications at the edge of the network and reduce system resource use. This can be achieved through configuration optimization, hard and soft restrictions, using a reduced container image, monitoring and adjusting container resources, and using lightweight base images. These measures help maintain system resiliency, improve system scaling decisions, and remove redundant resources.

Another approach is to use sequential downloads. Literature shows that parallel downloads are usually prone to blocking whereas sequential downloads allow for speedier image builds. Along with other techniques, like dependency management, as shown in the table, these measures substantially lowered the amount of disk usage for images.

Uninvestigated parts. To further improve the proposed methods, a further look can be taken at AI model optimization. Using technqiues like pruning, AI model can be fitted to a smaller device, like Raspberry PI nano. For a more robust testing another metric of resource utilization can be utilized, especially for specific optimization techniques. These approaches contribute to refining the performance and applicability of AI models across various platforms and scenarios.

Another area of research may be utilizing different devices and making load test devices. This research can provide additional information by assessing how various hardware configurations and environments influence the performance and scalability of systems or applications under different loads. By making experiments across a range of devices, we can gain insights into optimizing performance, resource allocation, and overall system efficiency across varied technological contexts. This exploration contributes to advancing the understanding and implementation of robust, adaptable systems capable of meeting diverse operational demands.

DEEP LEARNING METHOD FOR IDENTIFYING PROPAGANDA IN TEXTUAL DATA

Polina Shakhova — 2024-09-28

The paper examines the application of deep learning techniques for the detection of propaganda in textual data. The effectiveness and limitations of various neural network architectures, such as Long Short-Term Memory (LSTM) networks, Convolutional Neural Networks (CNNs), and Transformers, are analyzed. Particular emphasis is placed on the incorporation of hierarchical attention mechanisms to enable a more comprehensive understanding of propaganda techniques across different levels of text, including word, sentence, and document levels.

ACCELERATING ETL/ELT PROCESSES OF VOICE SIGNALS WITH GPU CLUSTERS: OVERVIEW

Andrii Didus — 2024-09-28

The implementation of GPU clusters in ETL (Extract, Transform, Load) and ELT (Extract, Load, Transform) processes addresses significant challenges in data-intensive tasks, including high-speed data processing, efficient resource utilization, and scalability. This paper overviews several approaches—RAPIDS Accelerator for Apache Spark, RAPIDS and Dask integration, Cloudera Data Platform with NVIDIA GPUs, and Cylon framework—highlighting their effectiveness in accelerating ETL/ELT operations for voice signals. The analysis provides insights into the advantages and limitations of each method, offering guidance for optimizing data transformation and analytics on GPU clusters.

SOME ASPECTS OF ONDRAW-ONTOUCH PATTERN IMPLEMENTATION FOR MULTIMODE SOFTWARE APPLICATIONS

Viktor Poriev — 2024-09-28

The article considers the specifics of the implementation of the onDraw-onTouch pattern for the organization of the graphical user interface on the example of building a software application of a geoinformation system for mobile devices

АНАЛІЗАТОР НАСТРОЇВ У КОНФЕРЕНЦІЇ НА ОСНОВІ WebRTC

Dmytro Tymochko — 2024-09-28

The article examines the feasibility of using a sentiment analyzer based on WebRTC to improve communication efficiency in conferences. The advantages and challenges of integrating a sentiment analyzer are considered, as well as the impact of this technology on the quality of interaction in distributed and centralized video conferencing systems. Examples of using a sentiment analyzer to enhance communication productivity in various fields are provided.

COMBINING PRETTY GOOD PRIVACY AND ROLE-BASED ACCESS CONTROL TECHNOLOGIES FOR ACCESS PROTECTION TO CONFIDENTIAL DATA

Danil Kolmahin — 2024-09-28

This article discusses a granular access protection model based on the combination of PGP (Pretty Good Privacy) and RBAC (Role-Based Access Control). The proposed model ensures an increased level of security through data encryption and role-based access control, enabling effective management of confidential data in modern information systems.

METHOD OF FAULT TOLERANT ROUTING IN DISTRIBUTED SYSTEMS BASED ON NON-BINARY DE BRUJIN TOPOLOGY

Oleksandr Honcharenko — 2024-09-28

The article considers a new method of routing in a system based on de Brujin topology, which is based on alphabetical decomposition and routing trees. Advantages and disadvantages compared to methods based on elementary transformations and based on simple decomposition into trees are considered

Ivan Shkardybarda, Artem Volokyta METHOD OF INCREASING THE FAULT TOLERANCE OF DISTRIBUTED DATA STORAGE SYSTEMS

Ivan Shkardybarda — 2024-09-28

The article addresses the issue of enhancing the fault tolerance of distributed data storage systems. For this purpose, the fault-tolerant de-Bruijn topology and the Raft consensus algorithm are utilized. These tools achieve high data consistency in not-fully connected topologies.

SOME METHODOLOGICAL AND ALGORITHMICAL ASPECTS OF TOPOLOGY ANALYSIS OF NETWORK STRUCTURES

Viktor Poriev — 2024-09-28

The article considers the methodological and algorithmic aspects and features of the implementation of software tools for the analysis of topologies of network structures, aimed at building high-performance and fault-tolerant computer systems.

WEB APPLICATION WITH TEXT MESSAGE ENCRYPTION FOR SECURE COMMUNICATION

Yaroslava Nehrych — 2024-09-28

In the article, the development of a secure web application for messaging using modern cryptographic methods is discussed. Particular attention is paid to the implementation of a fake password feature, which provides an additional level of protection for users in situations of coercion and extortion. Comparisons with existing messengers are considered, and how they implement additional security features is analyzed. Other extortion protection methods used in modern communication tools are also explored, and their advantages and disadvantages are evaluated.

METHOD OF DESIGNING A COMPUTER NETWORK ACCESS CONTROL SYSTEM

Mukhailo Valigura — 2024-09-28

This article presents a method to optimize access control systems in multi-tenant environments using Role-Based Access Control (RBAC) and an Access Request Optimization Algorithm (AOA). By integrating AOA, the system achieves improved performance and resource utilization. Experimental validation demonstrates significant enhancements in processing time and overall system productivity.

SYSTEM FOR DETECTING SCAM TOKENS AND SCAM TRANSACTIONS IN THE BLOCKCHAIN NETWORK

Oleh Lysenko — 2024-09-28

The article deals with the issue of cryptocurrency fraud, which affects more and more people. The main idea is to develop an analyser of new crypto projects with modern analysis methods and an intuitive interface, which is planned to be launched in the future.

Hrona, Klymenko. Development and production of LED wi-fi lamp

Yurii Hrona — 2024-09-28

The article considered the feasibility of creating an LED Wi-Fi lamp. Advantages and disadvantages of existing options on the market are considered. The process of creating the device and software for it is shown in detail.

METHOD FOR INCREASING THE EFFICIENCY OF SOLVING SLE IN A REDUNDANT NUMBER SYSTEM ON FPGA

Illya Verbovskyi — 2024-09-28

The article considers the possibility of increasing the efficiency of solving systems of linear equations, with reducing the number of necessary resources for data transfer on FPGA, due to the use of modules of digit-by-digit processing in an online mode with a redundant number system.

USING THE FLUTTER FRAMEWORK AS A MOBILE APPLICATION DEVELOPMENT TOOL THAT CAN AUTOMATE PROCESSES BETWEEN FINANCIAL AND TIME ASSISTANCE

Сергій Богаченко — 2024-09-28

The article considers the development of a mobile application for controlling financial expenses and free time of the user using the Flutter framework as a tool for implementation.

A PARALLEL METHOD FOR THE SWARM OF DRONE FLIGHT SIMULATIONS

Микола Ніколаєв — 2024-09-28

This paper introduces a novel method for the parallel simulation of drone swarms using shared-memory computing, enhancing efficiency and scalability. It details asynchronous execution strategies that optimize computational resources and demonstrate substantial performance gains in UAV coordination tasks.

METHOD OF IMPROVING THE EFFICIENCY OF VOICE INPUT FOR WRITING SOFTWARE CODE

Anastasiia Danevych — 2024-09-28

The article discusses the issue of writing computer code using voice input, issuing specific commands by voice. Modern methods of the stated task are examined, significant drawbacks of current solutions are identified, namely, the low level of context orientation. A solution to the identified drawback is proposed and the effectiveness of the new method is verified.

CPU CACHE SHARING DETECTOR FOR C/C++ PROGRAMS

Mykola Hohsadze — 2024-09-28

The paper discusses the false-sharing detector and the SHERIFF-DETECT detector, as well as its modification to create a more modern tool that can be used, more flexible in terms of architectural dependencies and using modern processor instructions. The advantages and disadvantages of different false-sharing detector approaches are considered.

A SOFTWARE SYSTEM FOR THE IMPLEMENTATION OF EXCURSIONS USING AUGMENTED REALITY AND ARTIFICIAL INTELLIGENCE

Danylo Monakov — 2024-09-28

The article examines the software system for the implementation of excursions using augmented reality and artificial intelligence. A method of conducting excursion activities based on the use of the latest AR and AI technologies is proposed.
The proposed system provides an opportunity to significantly improve the quality of excursion services thanks to the integration of augmented reality and artificial intelligence. The developed system can be effectively used in educational and cultural programs, providing interactive learning and involvement in cultural heritage.

DEVELOPMENT OF A SCALABLE AI PLATFORM BASED ON INTEGRATION OF EDGE COMPUTING WITH CLOUD TECHNOLOGIES

Volodymyr Rusinov — 2024-09-28

Relevance of the research topic. Recent trends show that AI is becoming more widespread in all areas of life. One such broad area is edge computing, and how to best leverage their strengths and weaknesses to deliver quality services to the end user. Today's cloud systems use a large number of data centers located around the globe to increase performance and improve latency, but they often do not offer flexibility in terms of user experience for both users and the companies that operate them.

In order to get an understanding of how much the system can be used as accurately as possible, we introduce several metrics that will help to obtain system parameters by which the proposed system can be compared with alternative solutions.

Formulation of the problem. The deployment of AI software should be accompanied by a thorough analysis of the target system. This article will explore different approaches to deploying AI in edge systems. One will include a smaller model suitable for low-power devices, and the other will be a full-scale model. Several tests will be conducted to measure the timing and accuracy of the proposed system.

Analysis of recent research and publications. Scientific research in areas related to increasing the speed and accuracy of delivery of AI-enabled systems is not widely explored. Edge and Fog computing technologies offer a wide range of performance optimizations, but little research has been published on running AI models on Edge/Fog platforms.

Setting objectives. The goal of the paper is to determine the feasibility of integrating edge computing with cloud computing and see how well it performs with AI tasks.

Presentation of the main material. Modern solutions, as a rule, involve the use of autonomous IoT devices [1] or cloud solutions [2]. Edge computing is a relatively new model of device networking, creating demand for solutions capable of rapid deployment and capacity scaling. Most research efforts are focused on building a more powerful AI model or scaling down the model to try to run it on an IoT device. Small current research includes using the cloud to facilitate the penetration of artificial intelligence and exploring ways to improve the quality of service. The resulting system can be applied in a variety of existing industries, in order to obtain significant economic and social benefits at a relatively low cost of implementation.

Conclusions. This research shows a way to improve the performance of embedded artificial intelligence on IoT devices using an Edge architecture. To illustrate this, a system was developed that demonstrated the ability to rapidly deliver ML results over a network. Using cloud computing resources, we were able to deploy such a system much faster than using a conventional on-premise cluster, but this approach is not limited to the cloud. There is room for improvement. The accuracy of DNNs can be improved by various means, for example, further tuning of the hyperparameters or changes to the architecture can bring benefits. Using RNNs can be useful for time-sensitive data, or other datasets can be used to improve accuracy by augmenting the original dataset. All this can be easily done in the cloud. This research proves that we can extend advanced computing capabilities with cloud services.

Unexplored aspects. The reliability of this solution can be related to its architecture. It is worth investigating the proposed method with different settings, for example: increasing the number of EC2 instances to ensure a larger data throughput, creating a separate service for data management. The regional infrastructure solution presented in this publication can be improved by adjusting the parameters and topological organization of the network. Perhaps more cloud services can be hosted on the client side to reduce costs. Finally, additional studies should be conducted to determine the financial benefits of the system obtained by this method.

Keywords: AI, clouds, scaling.

DYNAMIC RECONFIGURATION METHOD FOR SOFTWARE-DEFINED WIRELESS NETWORKS

Dmytro Oboznyi — 2024-09-28

The work considers the selection of typical geometric graphs for the generation of topological software and configuration support when using wireless communication methods. It is proven that currently the data transmission channel has the highest characteristics.

Enhancement of the dependency reduction method for microservice architectures

Bohdan Shmalko — 2024-09-28

The paper explores methods of reducing dependencies in microservice architectures to enhance their efficiency and stability. Innovative management techniques are developed and tested, demonstrating that these methods improve system robustness and scalability. Comparisons with existing solutions underline the unique advantages of the proposed approach, notably in cost-efficiency and ease of integration, making it suitable for modern technology needs.

INCREASING THE DATA TRANSMISSION SPEED CONSIDERING QOS PARAMETERS IN SDN NETWORKS UNDER THE ONOS CONTROLLER MANAGEMENT

Oleksii Cherevatenko — 2024-09-28

The article considers the possibility of implementing multiple analysis of quality of service parameters during traffic transmission in an SDN network under the control of the ONOS controller. It has been confirmed that considering the quality of service of communication channels during data transmission in an SDN network leads to the preservation of data transmission speed when some of the network nodes and channels are disconnected.