Volume 16 - 2024
Analysis and Testing of Next-Generation Passive Optical Access Network
In the past decade, we have witnessed the rapid development of the Internet as well as corporate networks. The evolution of mobile Internet, global digitalization, cloud computing, and the Internet of Things has consistently increased the bandwidth requirements across all segments of computing systems, from access networks to core networks. Under these conditions, optical access networks have emerged as an optimal solution for the implementation and further development of the access segment due to their service transparency, economic efficiency, energy savings, and higher security compared to other access networks. The International Telecommunication Union (ITU), together with the Full-Service Access Network (FSAN) organization, has established the necessary development steps and standards for next-generation access networks to meet the increasing capacity demands. Passive optical networks (PON) utilize passive components with low energy requirements, eliminating the need for power in the optical fiber distribution network. PON is currently primarily deployed in optical access networks as a Fiber to the Home (FTTH) solution, enabling triple-play broadband services of voice, data, and video over a single optical fiber shared among multiple users in residential areas. Managing these networks requires adequate tools for efficient cost monitoring. This includes troubleshooting with the ability to remotely differentiate between equipment failures and fiber breaks, followed by network localization, which is critical for telecommunications operators. The use of Optical Time-Domain Reflectometer (OTDR) techniques is widespread in point-to-point optical network topologies.
The objective of this master’s thesis is to provide a comprehensive overview of the state of passive optical networks, exploring both theoretical concepts and practical applications. On the one hand, this work presents a complete and systematic review of passive optical access networks, addressing current research topics as well as key operational issues related to the guidelines for designing and implementing passive optical network architectures. Additionally, testing equipment for passive optical networks is presented, with detailed analysis of optical power and its measurement using optical power meters. On the other hand, the thesis analyzes a case study of FTTH network implementation in Serbia, encompassing conceptual design, adopted technologies, protocols, acceptance testing procedures, and service commissioning. The study includes methodologies employed by designers to calculate power budgets and transmission efficiency concerning required distances, operating wavelengths, split ratios, and minimum transmission power requirements. A step-by-step testing procedure is outlined to verify compliance with ITU standards and equipment specifications from various manufacturers. The results of OTDR measurements are graphically presented and thoroughly analyzed to verify the network's compliance with relevant standards and specifications.
Techno – Economic Analysis of Optical Fronthaul Network Implementation in 5G-Systems
In the rapidly expanding world of the Internet of Things (IoT), the number of devices connected to IP networks is expected to surpass the global population by the end of 2024. This growth is driven by advancements in mobile communication, particularly 5G technology. The increasing device connectivity has led to more RRU (Remote Radio Head) antenna sites and higher traffic volumes requiring efficient transmission to centralized BBU (BaseBand Unit) units. However, the fronthaul network connecting RRU and BBU units faces challenges due to traffic volume, connection density, and quality-of-service demands. Passive Optical Networks (PON) have emerged as a promising solution for these issues, offering cost-effective components, high link capacity, and low latency, making them ideal for integration with 5G systems.
This paper explores the significance of 5G for IoT development and examines optical fronthaul networks in this context. It analyzes radio access network (RAN) architectures, including centralized RAN (CRAN), and reviews fronthaul systems and TWDM PON networks within 5G architectures. The thesis also focuses on minimizing the Total Cost of Ownership (TCO) in 5G implementations by optimizing capital (CapEx) and operational expenditures (OpEx). A mathematical model using the Integer Linear Program (ILP) approach is proposed, along with two heuristic algorithms: K-means clustering and Genetic Algorithm (GA). These algorithms are evaluated through a case study, incorporating real-world implementation parameters. Results are presented with diagrams, tables, and figures, offering insights into the methodology's practical implications.
Implementation and comparison of different reverse proxy and load balancing methods and solutions
This paper explores the implementation and comparison of reverse proxies and load balancing methods across Layer 4 (L4) and Layer 7 (L7) of the OSI model. Reverse proxies and load balancers are essential components in modern distributed systems, ensuring scalability, reliability, and performance optimization. The research focuses on evaluating the performance of custom L4 and L7 reverse proxy and load balancing solutions against popular open-source tools such as NGINX and Traefik.
This paper contributes to the understanding of reverse proxy and load balancing mechanisms, presenting insights into their application in high-performance systems. Recommendations for further research include optimizing TLS handling, improving multi-threaded processing, and exploring dynamic load balancing algorithms for distributed architectures. The findings are valuable for designing scalable and efficient network infrastructure in modern software systems.
Authentication Methods of IoT Devices in Communication With the Server: Implementation and Comparative Analysis
The Internet of Things (IoT) is revolutionizing everyday life and industrial processes by enabling connected devices to communicate, analyze, and act autonomously. As IoT devices increasingly transmit sensitive data, secure communication becomes paramount, with authentication serving as the foundation for ensuring data confidentiality and integrity. This paper explores various authentication methods tailored to IoT devices communicating with servers via the HTTP protocol, offering theoretical insights and practical implementation guidelines.
Each method is thoroughly examined concerning its advantages, disadvantages, and suitability for different IoT environments. Implementations are provided for both client-side (IoT devices) and server-side using technologies like JavaScript and PHP (Laravel). Particular attention is given to the resource constraints typical of IoT devices, such as limited processing power and memory.
Application of Artificial Intelligence in Predictive Maintenance of Urban Infrastructure
This paper explores the application of artificial intelligence in predictive maintenance of urban infrastructure, with a particular focus on bridges. A machine learning model was trained using simulated data to detect anomalies, analyzing key factors such as temperature, load, and vibrations. The evaluation demonstrated high model accuracy, highlighting its potential for practical implementation. The discussion addresses integration with IoT sensors and scalability in real-world scenarios, providing recommendations for further development. Challenges related to the availability of real-world data and the need for adaptable solutions are also analyzed, laying the groundwork for future research and practical applications.
Holistic Knowledge Representation and Structure Learning
This paper addresses the challenge of holistic knowledge representation and structural learning through the concept of structured representation in one-dimensional spaces. Unlike traditional natural language processing (NLP) techniques, which focus on distributed representations and partial consideration of semantics, the proposed approach enables comprehensive text analysis by integrating hierarchical data organization.
The proposed method involves creating dynamic structures for representing textual corpora, where each level of the structure is based on unique objects that represent elementary and complex units, such as words, phrases, sentences, and paragraphs. By employing algorithms for object recognition, creation, and reorganization, the method facilitates efficient management of large volumes of textual data.
The results demonstrate that the proposed model overcomes the limitations of existing methods, such as scalability and precision in text analysis, while maintaining flexibility in modeling new information. The conclusions highlight the significant potential of structured representation in tasks such as text classification, sentiment analysis, and language generation.
Communication patterns in microservice architecture
The paper focuses on the exploration of communication patterns in microservice architecture, with special emphasis on technologies that enable seamless interaction among distributed services. In addition to isolating functionalities, microservices require efficient orchestration and coordination in order to collectively fulfill complex business processes. In this sense, communication mechanisms play a key role in ensuring the coherence of the entire system. The aim of this research is to provide a comprehensive overview of different communication patterns within the context of microservice architecture. This paper covers synchronous communication through simple REST APIs, followed by asynchronous communication using RabbitMQ as a message broker.
Demand Response
In today's world, electrical energy serves as the foundation of everyday life and industrial growth. With the emergence of modern technologies, traditional power grids have evolved into smart grids, designed to meet the growing demand for efficient and reliable energy sources. This paper explores the evolution of these grids, analyzes the current state of power grid infrastructure, and examines the challenges and opportunities presented by smart grids. In the digital age, where sustainability is key, smart grids play a crucial role in modernizing the energy sector, impacting both technological progress and socio-economic dynamics. A significant part of this modernization is Demand Response (DR), a strategy that enables consumers to adjust their electricity consumption during peak demand periods, offering both economic and environmental benefits.
Developing Mobile Applications Using Cloud Solutions: Case Study of the RAF Network Mobile Application
This paper explores the concept of developing mobile applications using cloud solutions, focusing on the RAF Network mobile application. Through a detailed analysis of the technologies and methodologies employed in the development and implementation of this application, the paper demonstrates how cloud computing can significantly enhance the functionality, scalability, and reliability of modern mobile applications. The RAF Network application was developed using the Flutter framework for the frontend and the Django web framework for the backend, with cloud infrastructure hosted on the DigitalOcean platform. The application enables seamless communication between the client and server sides through a REST API exposed via a Traefik reverse proxy. This approach ensures high scalability and flexibility, allowing for the easy addition of features such as integration with student services and enhanced communication between students and professors. Additionally, Firebase, a cloud-based platform developed by Google, was integrated to provide comprehensive tools for mobile app analytics, performance monitoring, and real-time error tracking. This enhances the app's ability to gather critical insights and optimize user experience. The paper highlights the advantages of using cloud technologies in mobile application development and proposes future improvements in functionality and user experience.
Development of a Programming Learning Platform - Server Side
Educational institutions play a crucial role in equipping students with programming skills, making programming education a key segment of their curricula. To facilitate efficient learning and assessment, institutions require robust and adaptable platforms that support programming tasks and exams, ensure seamless integration with development environments, support collaborative learning, and provide secure and scalable operations. This paper addresses the challenges faced by existing programming learning platforms, such as limited integration, insufficient support for collaborative tasks, and lack of adequate security measures. It presents a solution involving the development and implementation of a robust server-side infrastructure for a programming learning platform. The platform integrates with IntelliJ IDEA, provides a private Git server, and utilizes a REST API to support various educational functions. The architecture, design, and key components of the system, including the database model and security measures, are discussed. The paper also highlights the benefits of an agile development methodology and presents plans for future enhancements, including developing extensions for IntelliJ IDEA and VS Code for teachers and students.