06.943GPPTSVoice Activity Detector (VAD) for Adaptive Multi Rate (AMR) speech traffic channels
The instantaneous frame SNR, SNR, and long-term peak SNR, SNRp(m), are used to calibrate the responsiveness of the VAD decision. When the frame count is less than or equal to four (m 4) or the forced update flag (sec 4.3.10) is set (fupdate_flag == TRUE), then the SNR’s are initialized as:
Otherwise, the instantaneous frame SNR is generated by:
and the long-term peak SNR is derived by the following expression:
The long-term peak SNR is then quantized in 3 dB steps and limited to be between 0 and 19, as follows:
where is the largest integer x (floor function).