作者: K Sabnani , JC Lin , S Paul , S Bhattacharyya
DOI: 10.1109/49.564138
关键词: Packet switching 、 Computer network 、 Latency (audio) 、 Network packet 、 Computer science 、 Real-time computing 、 Reliable multicast 、 Selective Repeat ARQ 、 Throughput 、 Scalable Reliable Multicast
摘要: This paper presents the design, implementation, and performance of a reliable multicast transport protocol (RMTP). The RMTP is based on hierarchical structure in which receivers are grouped into local regions or domains each domain there special receiver called designated (DR) responsible for sending acknowledgments periodically to sender, processing acknowledgment from its domain, retransmitting lost packets corresponding receivers. Since recovered by retransmissions as opposed original end-to-end latency significantly reduced, overall throughput improved well. Also, since only DRs send their instead all single generated per region, this prevents implosion. Receivers periodically, thereby simplifying error recovery. In addition, selective repeat retransmissions, leading at cost minimal additional buffering also describes implementation Internet.