4.2 GENERIC VAD/CNG FUNCTIONALITY
Generic VAD/CNG functionality is illustrated in Fig. 4.1(a) for the G.729AB encoder and decoder. For simplification of representation, only the encoder and decoder are shown in the voice call, and voice is shown in one direction in Fig. 4.1(a). G.729AB has a built-in VAD/CNG function. In the VAD disable mode, G.729AB generates voice packets for every 10 ms or multiples of 10 ms. The packets are marked on the top row as “S S S” in boxes. In the VAD-enabled mode, speech packets marked as “S” and the VAD–SID packet marked as “V” are sent on the network. The packets shown without any marking (empty boxes), are of no packet or a null packet. For completeness of representation, Fig. 4.1(a) and (b) are shown with an empty box for no packets. The empty boxes are the bandwidth-saving operations of not sending any packet on the network. At the destination, the decoder decompresses speech and VAD packets. The VAD packet helps in continuing the generation of comfortable noise during no packets.
The G.711/726-based VAD/CNG is shown in Fig. 4.1(b). In this scheme, external VAD and CNG functions are required. For simplification, the VAD, SID packet, and DTX are referred to as VAD. The VAD module generates detection of a decision regarding speech or silence and a decision regarding when to send an updated packet, which is controlled by the DTX operation. The VAD module provides a required selection to choose between speech or VAD packets. With reference to Fig. ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access