20.6 SUMMARY AND DISCUSSIONS

VoIP users are accustomed to the high quality offered by traditional PSTN service. In many situations, VoIP does not offer consistent voice quality. Voice quality in VoIP networks widely varies as a result of several factors. This chapter is presented on voice quality evaluation and how to avoid service degradations in a VoIP network and how to deliver voice quality that is comparable with, or even exceeds, PSTN levels. In this chapter, voice quality influencing parameters and their improvements are presented. VoIP systems have to cater to multiple countries and multiple customer deployments. QoS mechanisms and monitoring are also required to be incorporated. The goal in narrowband (300 to 3400 Hz) voice quality is to meet PSTN quality. In the wideband case, the perceptions have to exceed PSTN reference quality. Several options for measuring voice quality are available. The passive E-model can be defined as a function of room noise, end equipment impairments, echo, delay, codec, and packet loss. PESQ is a function of coding distortions, errors, noise, filtering, and variable delay. P.563 is a function of noise, filtering, and variable delay as well as distortions caused by channel errors and speech coders. The first level of quality is to ensure by design. Measurements and monitoring will also help in giving feedback. In addition to these several perception-based tests, user experiences in different conditions are also important.

VoIP Voice and Fax ...

Get VoIP Voice and Fax Signal Processing now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.