Speech Coding In 3G Systems

The tremendous success of second generation (2G) cellular telephony was largely enabled by use of speech compression techniques that provided network capacity benefits and decent voice quality across all wireless link characteristics, such as fading and transmission errors, by means of advanced digital error correcting techniques. With an increasing demand on voice quality—comparable to or better than PSTN—and additional new services that also require large data bandwidth, such as video and still image, advanced speech compression will continue to play an important role in 3G networks. This paper gives an overview of speech coding standards designed for those wireless systems, and discusses the three directions considered by 3G wireless networks:

1. Use a variable low bit-rate codec to improve voice quality and network bandwidth efficiency (e.g., IS-893, SMV)
2. Use the same speech codec in 2G/2.5G in order to allow TFO with 3G system (e.g.,GSM AMR = AMR-NB)
3. Use more data bandwidth to provide wider frequency band voice services (e.g. AMR-WB, 7 KHz vs. PSTN's 3.4 KHz).