Voice Flow Regulator in a Digital Packet Voice Network
Original Publication Date: 1985-Mar-01
Included in the Prior Art Database: 2005-Feb-18
In a digital packet voice network, voice packets are created by a speech analyzer located at the emitter, and transmitted at regular time intervals (say 20 ms) over the network, but are received at randomly distributed time intervals at the receiver. This time jitter is mainly due to queuing delays within the network nodes. Since on the one hand voice packets must be provided to the synthesizer located at the receiver at regular time intervals so as to meet the realtime constraint of speech, it is necessary to implement a re-synchronization by means of a classical packet queue with retention delay. Since on the other hand the maximum inter packet time can vary during the day and depends on the route to be used, the retention delay should theoretically be set to the maximum expected delay within the network.