Hybrid model of monitoring computer network metrics

DOI: 10.31673/2412-9070.2025.042533

Abstract

The article examines the problem of obtaining and analyzing key computer network metrics, such as bandwidth, latency, jitter, and packet loss, with a particular focus on enhancing the accuracy, adaptability, and responsiveness of monitoring systems. In the modern context of digitalization, where network traffic is constantly growing and network structures are becoming increasingly complex, the timely detection of changes and potential issues in network infrastructure operation is a critically important task.
In this regard, the paper proposes the use of a specialized hybrid monitoring model that integrates the capabilities of Software-Defined Networking (SDN) and Network Function Virtualization (NFV) with fuzzy logic inference mechanisms. Such integration enables the creation of a flexible and scalable solution capable of automatically responding to load changes and ensuring high levels of network service performance, stability, and reliability.
The rationale for developing such an adaptive model lies in the fact that traditional monitoring tools based on static algorithms and limited data collection protocols often fail to adequately reflect the real-time state of the network. The study demonstrates that flexible approaches using fuzzy logic control allow for automation in decision-making related to load balancing, routing, traffic priorityza tion, and resource scaling.
The research presents a conceptual model for acquiring key metrics, which includes phases of active and passive monitoring, as well as traffic segmentation by service class. Additionally, automatic adjustment of virtualized network function parameters is provided based on analytics obtained through fuzzy logic. This approach ensures more efficient allocation of network resources, improves quality of service for end users, and guarantees system adaptability to sudden traffic spikes. It signify cantly reduces the risks of overload and packet loss, which are critical for maintaining uninterrupted operation of digital services—especially in real-time applications such as VoIP, video conferencing, and streaming.

Keywords: computer networks; monitoring methods; throughput; SDN/NFV; fuzzy logic inference; load; data packet; traffic; adaptive model; performance.