Design and Implementation of a Hybrid Framework for QoS Monitoring in 5G-Enabled Services

Abstract

The fast development of 5G networks presents complicated Quality of the Service (QoS)-based management issues that cannot be managed using the conventional monitoring methods. In this paper, a new framework being proposed is a hybrid system between active probing and passive monitoring as a way of having a full evaluation of the QoS in 5G-enabled services. The proposed system will integrate real-time feedback with AI-driven analytics by using standardised KPIs throughput, latency, jitter, and packet loss. Open5GS and Wireshark (with a machine learning module to detect anomalies) were used as open-source platforms to implement the architecture. The experiments on an emulated 5G core and the comparison with single-mode methods showed an accuracy of measurement improvement by 27 %. The system has a scalable deployment, dynamic sampling, and visualization of network dynamics without interruptions to be made in the diversity of traffic. The hybrid model allows the assurance of constantly changing QoS of new applications, such as autonomous vehicles, industrial IoT, and smart city infrastructure.