Abstract

The ability to communicate wirelessly in different domains is about to change irreversibly with the arrival of 5G technology. Mudering 5G diverse verticals, however, including ultra reliable low latency communication (URLLC) and massive machine-type communications (MTC), will require constant and proper monitoring of the networks so appropriate levels of Quality of Service (quality of service) can be guaranteed. This paper explores the most recent tools to measure and evaluate the QoS in a 5G system \- performance tailoring to the specific needs of the 5G network. The paper briefly assesses active and passive measurement tools, including iPerf, NetPerf, Wireshark, Open5GS, and sets of 5G QoS Probes, primarily focusing on throughput, latency, jitter, and packet loss. The paper also touches on the phenomenon of measuring QoS in 5G networks' heterogeneous architecture (HetNet) and the specific challenges posed by dynamic slicing in MEC and Network Function Virtualization (NFV) with QoS. The paper also discusses real-time monitoring and QoS evaluation via Anomaly Detection, as well as QoS prediction using Artificial Intelligence and machine learning. For the analysis above, the paper discusses several use cases of smart cities, autonomous driving, and telemedicine while also providing some of the comparative studies of the tools in question, stressing the absence of standardization and interoperability of the measuring tools, which causes the lack of reliable and consistent performance evaluation of the networks in question. This work evaluates existing systems and provides a basis for building more sophisticated and responsive quality of service monitoring systems for 5G telecommunications.