Joint capacity and quality evaluation for AMR telephony speech in WCDMA systems

The adaptive multi-rate (AMR) speech codec is the mandatory speech codec for WCDMA systems. The codec supports eight different source rates ranging from 12.2 kbit/s down to 4.75 kbit/s. This paper evaluates different ways of allocating AMR rates to users on the downlink in a WCDMA system. A novel system performance concept is introduced based on a user satisfaction metric that tries to combine the user experience from different speech quality levels as well as events like blocking and dropping. System simulations show that the AMR codec introduces a significant trade-off between capacity and quality for the speech service. By allocating AMR modes based on the system load, the quality and capacity trade-off can be efficiently balanced and high system performance can be achieved for a wide range of offered loads.