A qualidade da voz em sistemas de telecomunicações

Abstract

This work compares the main methods for measuring the quality of voice in telecommunications systems. Initially, the processes of hearing and speech production are described, as well as physical and organic factors that contribute to good communication. Then, the concepts related to the needed coding to ensure the voice transmission in a telecommunications network are discussed. The types of transmission, the coding process and the coders are presented, as well as the conventional telephone networks and the VoIP. In order to better understand the methods of measuring quality, the parameters involved in the computation of the numerical values related to the voice quality are described: the organic and physical parameters (pitch, loudness, etc.), and specific parameters of telecommunication networks (jitter, latency, etc.). The measurement methods are then presented, divided into two classes: subjective and objective methods. Among the objective methods several options are presented, separated into three method groups: the signal-based intrusive, the signal-based non-intrusive and the parameter-based. Finally, experiments are presented in TDM and VoIP environments with the PESQ and PAMS signal-based intrusive methods using the Performer equipment from RADCOM, available at CRT (EMBRATEL). A comparative analysis of the results is performed, including the E-model, which is a network parameter-based method. Conclusions about the advantages and disadvantages of each method used in the experiments to evaluate the transmitted voice quality are presented. A detailed instruction of the Performer equipment is also inserted in this work.