Parametric Models for Speech Quality Estimation in GSM Networks

Speech quality assessment is an emergent QoS aspect in cellular network monitoring. This contribution deals with the estimation of perceived speech quality in GSM networks, by means of parametric models. Our goal is to provide a model for MOS estimation, based on parameters available in the operation and maintenance centre (OMC) and in the measurement reports from mobile stations. Specifically, based on MOS values from a large database obtained performing intrusive tests in a real GSM network, and on the correspondent measured transmission parameters, an extensive data analysis has been conducted. The correlation coefficients of GSM parameters with the objective speech quality have been then maximized, and a simple linear model has been proposed. Moreover, handover events have been considered and their impact on perceived speech quality has been proved