Intelligent voice smoother for silence-suppressed voice over Internet

For transporting voice data with silence suppression over the Internet, the problem of jitter introduced from the network often renders the speech unintelligible. It is thus essential to offer intramedia synchronization to remove jitter while retaining minimal playout delay. We have previously proposed a neural– network–based intra–voice synchronization mechanism, called Intelligent Voice Smoother (IVoS). IVoS is aimed to furnish CBR playout during talkspurts assuring minimal mean and variance of Distortion of Talkspurts (DOT) and mean playout Delay (PD). To substantiate the viability of IVoS, in this paper we present a prototyping system. The prototyping system is composed of the Silence Suppressor and Traffic Simulator modules at the sending host, an IVoS module at the receiving host, and a stand–alone Ethernet connecting the two hosts. The Silence Suppressor module caries out silence suppression based on two thresholds. The Traffic Simulator module is responsible for the creation of delay and jitter. The IVoS module simply performs major IVoS functions. The prototyping results reveal that, compared to existing playout approaches, IVoS achieves superior playout yielding negligible DOT and PD irrespective of traffic variation.