ERROR AND FLOW-CONTROL PERFORMANCE OF A HIGH-SPEED PROTOCOL

Recently, Netravali, Roome, and Sabnani proposed a lightweight protocol (called SNR) for the transport layer of a high speed data communications network. The key ideas of this protocol are periodic exchange of the full state of the two communicating entities, reduction of the protocol overhead by making decisions about blocks of packets rather than individual ones, and the ability to parallelize protocol processing. This paper studies the performance of the SNR protocol when implemented for end-to-end flow and error control. Using a combination of analysis and simulation, we evaluate the efficiency with which this protocol uses the network bandwidth and its achievable throughput as a function of certain network and protocol parameters, such as bandwidth-delay product, packet loss rate, the roundtrip delay, period of state exchange, the block size, correlation in errors/losses, etc. In addition, we suggest an enhancement of the protocol by introducing two windows to decouple the two functions of receiver flow control and network congestion control. This enhancement, as well as the original protocol, are compared with Go-back-N (GBN) and One-at-a-time-Selective-Repeat (OSR) retransmission procedures, and are shown to have significantly higher throughput for a wide range of network conditions. As an example, for a virtual circuit with 60 ms roundtrip delay and 10(-8) bit error rate, in order to deliver 500 Mbits/s throughput, both the GBN and OSR require a raw transmission bandwidth of approximately 800 Mbits/s; whereas SNR with two windows needs slightly higher than 500 Mbits/s raw bandwidth. Periodic exchange of state can also provide a variety of measures for congestion control in a timely and accurate fashion. This will be discussed in a forthcoming paper.