A selective repeat ARQ scheme for point-to-multipoint communications and its throughput analysis
SIGCOMM '86 Proceedings of the ACM SIGCOMM conference on Communications architectures & protocols
NETBLT: a high throughput transport protocol
SIGCOMM '87 Proceedings of the ACM workshop on Frontiers in computer communications technology
Multicast routing in datagram internetworks and extended LANs
ACM Transactions on Computer Systems (TOCS)
ACM SIGCOMM Computer Communication Review
A comparison of sender-initiated and receiver-initiated reliable multicast protocols
SIGMETRICS '94 Proceedings of the 1994 ACM SIGMETRICS conference on Measurement and modeling of computer systems
Distributed process groups in the V Kernel
ACM Transactions on Computer Systems (TOCS)
vic: a flexible framework for packet video
Proceedings of the third ACM international conference on Multimedia
A reliable multicast framework for light-weight sessions and application level framing
SIGCOMM '95 Proceedings of the conference on Applications, technologies, architectures, and protocols for computer communication
A bandwidth analysis of reliable multicast transport protocols
COMM '00 Proceedings of NGC 2000 on Networked group communication
Optimal branching factor for tree-based reliable multicast protocols
Computer Communications
Hi-index | 0.24 |
This paper analyzes the sender-based and receiver-based reliable multicast protocols for group communications and derives simple analytic formulae to characterize the bandwidth efficiency of the two techniques. The sender-based protocol requires all receivers to return positive acknowledgments (ACKs). This may lead to an ACK implosion problem. The receiver-based protocol requires the receivers to be individually responsible for detecting lost packets and raising repair requests for retransmission. This is more efficient, but requires more effort to implement. The analysis covers the following error recovery schemes: sender-based selective repeat, receiver-based selective repeat and receiver-based go-back-N. The results show that the receiver-based protocol is much more superior than the sender-based protocol. The receiver-based selective repeat performs better than the receiver-based go-back-N. The performance gain is particularly significant when the receiver-based protocol is enhanced with a technique to multicast repair request packets.