INPUT RATE FLOW-CONTROL FOR HIGH-SPEED COMMUNICATION-NETWORKS - BLOCKING AND DELAY AT THE ACCESS POINTS

We present queueing models to study the system performance at the access points of a high-speed communication network. An input rate control mechanism, using a Leaky Bucket type scheme (also identified as a Credit Manager Algorithm), is applied to regulate traffic from end-user stations to the network. For the performance analysis, we select a sub-network topology which consists of a network switch (such as a fast packet or cell switch in high-speed metropolitan or wide area networks) and a number of regulated source stations. To reduce packet or cell loss probabilities at the access switch, each user station implements locally a replica of the input regulation scheme. The output traffic streams from the source stations feed into the switch, which is modeled as a multiple-server queueing system. An efficient recursive method is introduced to characterize the output traffic streams from the source stations as Markov Modulated Processes. We demonstrate the ensuing tradeoffs between message delays incurred at source stations due to the input rate control mechanism and the message delays incurred at the switch as affected by the statistical multiplexing of the input message streams. Performance curves are presented to illustrate the statistical behavior of the queue-size and message delays. Using our results, the system designer can properly select the parameters of the flow control scheme to guarantee acceptable limits of queue sizes and message delays at the source stations and at the network switches.