Kang Min Gyu, Kung Sanghwan, Kim Dong Kyoo

The Transactions of the Korea Information Processing Society (1994 ~ 2000), Vol. 5, No. 10, pp. 2641-2653, Oct. 1998
10.3745/KIPSTE.1998.5.10.2641,   PDF Download:

Abstract

This paper addresses the problem found with audio packet loss on the teleconferencing system. In ITU H.323 teleconferencing system, the RTP/RTCP protocol offered to delivery realtime stream (audio and video). Participants in teleconferencing system have experience in packet loss which results in the congestion. The quality of the audio that delivered from source to destination depends essentially on the number of lost packets and on the delay variations between successive packets. We measure the loss of audio packets in order to examine the loss characteristics over Internet. The experience has shown that almost packet loss is isolate fairly over congested network. The goal in our study is to develop efficient mechanism to adapt the loss characteristics of the networks. By examining the packet loss of audio streams sent over Internet, we suggests that Dynamic Redundant Transmission mechanism would be adquate to minimize the impact of audio packet loss. The sending sites uses the feedback information which contains into the RTCP packets, to compute packet loss. Based on this packet loss state, the sender determines the policy of redundant transmission which is one of no-redundant, 1-order redundant, and 2-order redundant. These mechanisms are implemented and evaluated over the Internet. The results indicate that the mechanism can be applied to minimize almost all the non-consecutive packet loss and a part of consecutive loss.

Statistics
Cite this article