Pryslupskyi A. Improving the quality of experience of infocommunication services in the next generation intelligent networks

The dissertation solves the scientific and practical task of improving the quality of service perception in modern infocommunication systems by developing new methods of intelligent monitoring of the network status, allocation of network resources and quality of service management in the context of adaptation to changing user requirements and limited network resources. In the first chapter, “Analysis of existing methods and models for managing the quality of service perception in modern telecommunication networks”, the author reviews the architecture of software-configurable networks and highlights the main advantages of using them over traditional networks in terms of managing the quality of service provision. It is established that methods of ensuring the quality of service (QoS) are crucial for all organizations that want to guarantee the best quality of perception of their most important applications and services.In the second chapter, “Models and methods of building intelligent networks with adaptive resource management based on the quality of service perception”, a conceptual model of an intelligent intent-based network (IBN) is proposed, which is deployed on the basis of software-defeined networking (SDN) technology. According to the conceptual ideology, IBN offers network administrators a simple way to express business goals in the form of intentions, one of which is to ensure the required level of service quality, enabling network software to automatically achieve the set goals based on intelligent analysis of resource status and traffic management. This chapter proposes an intent-based QoE monitoring system for future software-defined networks that will improve the quality of service for end users and allow for more efficient use of network resources. In this section, we also developed a machine learning module for integration into software-defined networks. This allows predicting the level of quality of end-user service perception, taking into account network parameters such as delay and packet loss. A method for managing the quality of service perception in smart grids has been developed, which, unlike the known ones, is based on the intentions of users defined in the form of subjective QoE estimates to ensure the ordered quality of service. A modified method for migrating switches from one controller to another, taking into account the distribution according to QoE priorities, has been developed. In the third chapter, “Development of a unique fault-tolerant controller for client-oriented QoS management in next-generation software-configurable intelligent networks,” a unique IBN controller for intelligent software-defined networks was developed that provides customers with a reliable connection. This controller has an authorization function so that users can log in and use their account for all network manipulations. Also, the controller can be constantly improved and more and more new and useful functions can be added, which can develop in parallel with the development of the network itself. An automated system for restoring the availability of servers on which the SDN/IBN controller and IoT broker are deployed is proposed. The architecture of the server availability recovery system is developed. A system for monitoring the functioning of servers has been created, which makes it possible to increase the fault tolerance of the centralized smart grid management controller. For this purpose, a number of functioning algorithms have been developed, namely a block diagram of the Jenkins pipeline, remote server monitoring, and a remote server monitoring script. In the fourth chapter, “Practical implementation of an intelligent network based on the use of SDN ZODIAC technology and automation of the developed methods of managing the quality of service perception,” a module for intelligent control of the handover procedure based on the QoE parameter was developed for integration into wireless software-defined networks. Using the developed module, the handover procedure can be performed not only based on the signal strength of the access point, but also taking into account such network parameters as delay and packet loss. Taking these parameters into account allowed us to combine handover and dynamic QoE routing to ensure a high level of perception quality. According to the results obtained, the proposed algorithm allows us to quickly respond to sudden deteriorations in the network and provide the required quality of perception for the end user. For the practical implementation of the next-generation intelligent network, we used SDN Zodiac technology equipment, which, unlike proprietary network equipment manufacturers, is open for modification and allows software implementation of own resource management solutions.