A Simulation to Compute the Mean Opinion Score (MOS) for EVS-SW Audio Codec in 5G Mobile Networks

This article evaluates the Enhanced Voice Service SuperWideband (EVS-SW) audio codec performance throughout simulation scenarios developed in Python to compute the Mean Opinion Score (MOS) for 5G Mobile Networks. Different bit-rate value combinations corresponding to the SuperWideband (SW) interval were considered. Furthermore, the simulation scenario considers various separation distances between the User Equipment (UE) and the Base Station (BS). Different numbers of handover and speed are also evaluated to assess the influence on MOS and Quality of Experience (QoE). Using Machine Learning (ML) techniques, it also analyzes the independent predictors and an MOS-dependent variable to determine the actual influence of the audio codec. The findings indicate that bit-rate values do not significantly affect MOS and QoE; so do the separation distances and the number of handovers influence MOS. This work offers 5G mobile network operators (MNO) potential tools to identify and address issues swiftly, enhancing user experience without manual assessments.