ITU-T G.722.2 PDF

This Recommendation describes the detailed mapping from input blocks of 320 speech samples in 16-bit uniform PCM format to encoded blocks of 132, 177, 253, 285, 317, 365, 397, 461 and 477 bits and from encoded blocks of 132, 177, 253, 285, 317, 365, 397, 461 and 477 bits to output blocks of 320 reconstructed speech samples. The sampling rate is 16 000 samples/s leading to a bit rate for the encoded bit stream of 6.60, 8.85, 12.65, 14.25, 15.85, 18.25, 19.85, 23.05 or 23.85 kbit/s. The coding scheme for the multi-rate coding modes is the so-called Algebraic Code Excited Linear Prediction Coder, hereafter referred to as ACELP. The multi-rate wideband ACELP coder is referred to as AMR-WB. The codec described in this Recommendation also utilizes an integrated Voice Activity Detector (VAD).

The foreseen applications for this Recommendation are the following: Voice over IP (VoIP) and Internet applications, Mobile Communications, PSTN applications, ISDN wideband telephony, ISDN videotelephony and video-conferencing.

In addition to the algorithm specified in the main body of this Recommendation, Annexes A and B and Appendix I provide supplemental functionalities allowing interoperability with GSM and 3GPP wireless systems. These functionalities have originally been developed for these systems, but their use is not limited to mobile applications. Annexes D and E describe test vectors and frame structure respectively. These annexes may be implemented independently of this main body specification according to the different requirements of systems deploying the AMR-WB algorithm:

• Annex A describes comfort noise aspects for use of the AMR-WB algorithm in source-controlled rate operation. The implementation of Annex A is essential for interoperability with GSM and 3GPP wireless systems.

• Annex B describes the source-controlled rate operation for the AMR-WB algorithm. The implementation of Annex B is essential for interoperability with GSM and 3GPP wireless systems.

• Annex D describes the digital test sequences, which are a verification tool providing an indication of success in implementing the AMR-WB codec.

• Annex E describes the recommended frame structure for use with the different modes of operation for the AMR-WB algorithm.

• Appendix I describes an example solution for error concealment of erroneous or lost AMR-WB frames.

For better usability, the ANSI-C code with the low-level description of all these functionalities have been grouped into a single annex, Annex C. Should there be any discrepancy between the descriptions in any of the different parts of this Recommendation and the implementation of such descriptions in Annex C, the descriptions in Annex C shall prevail.

In clause 8 a specific reset procedure, called codec homing, is described. This is a useful feature for bringing the codec into a known initial state (e.g., for testing purposes). Clause 9 specifies the Voice Activity Detector (VAD) used in this codec as well as in the source controlled rate operation (DTX) specified in Annex B. Clause 10 provides information on minimum requirements for support of AMR-WB modes in the 3GPP speech telephony service.