C. Williamson, "Optimizing File Transfer Response Time Using Loss-Load Curves: A Simulation Study", to appear in Internetworking: Research and Experience, 1995.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... results with those of Brakmo et al. Researchers frequently simulate rather than conduct live experiments of new or tuned wide area network flow and congestion control algorithms sometimes because they are experimenting with service disciplines not yet implemented in deployed networks [20, 9, 10], because some practical aspects of the algorithm have yet to be worked out [18] and because live experiments must be conducted carefully to obtain statistically significant comparisons [2] This raises a quandary: neither packetby packet simulation nor live experiments can convince ....

Carey Williamson. Optimizing file transfer response time using the loss-load curve congestion control mechanism. ACM SIGCOMM 93, pages 117--126, September 93.

....based on randomlygenerated or trivial network topologies are so common, while rigorous analyses of how the results scale or how they might change with a different topology are so rare. The state of the art in network modeling includes: ffl Regular topologies, such as rings, trees and stars (e.g. [12, 44, 64]) E. Zegura, K. Calvert, and M. Donahoo are with the College of Computing, Georgia Institute of Technology. ffl Well known topologies, such as the ARPAnet or NSFnet backbone (e.g. 5, 54, 65] ffl Randomly generated topologies (e.g. 55, 58, 61] The limitations of each of these are ....

C. Williamson. Optimizing file transfer response time using the loss-load curve congestion control mechanism. In Proceedings of ACM SIGCOMM '93, pages 117--126, 1993.

....of how the results scale or how they can be applied to actual networks are so rare. A selected review of the literature indicates the use of network models to study a wide range of problems. The state of the art in modeling includes: ffl Regular topologies, such as rings, trees and stars (e.g. [3, 8, 14]) ffl Well known topologies, such as the ARPAnet or NSFnet backbone (e.g. 1, 9, 15] ffl Randomly generated topologies (e.g. 10, 11, 12] The limitations of each of these are obvious: wellknown and regular topologies reflect only parts of current or past real networks; random topologies ....

C. Williamson. Optimizing file transfer response time using the loss-load curve congestion control mechanism. In Proceedings of ACM SIGCOMM '93, pages 117--126, 1993.

....smoothing of these estimates uses an exponential averager with an a chosen a priori, instead of using a fuzzy predictor. Further, they appeal primarily to intuitive heuristics, and do not use a formal control theoretic model, to develop their control. Williamson has proposed the Loss Load scheme [60, 61] that uses the throughput versus loss curve to compute an optimal sending rate. This approach has numerous lacunae. It ignores system considerations, such as the the fact that monitoring each connection at each switch poses a rather considerable burden on the switch controller. Also, a source may ....

C. L. Williamson, Optimizing File Transfer Response Time Using the Loss-Load Congestion Control Mechanism, Proc. ACM SigComm 1993, 1993.

No context found.

C. Williamson, "Optimizing File Transfer Response Time Using Loss-Load Curves: A Simulation Study", to appear in Internetworking: Research and Experience, 1995.

....with a smaller number of packets to transmit, or with a longer round trip delay to the destination, may be willing to sacrifice slightly on raw throughput in order to minimize the excessive delay of end to end packet retransmission. For such a source, the response time model offers the answer [27]. The response time model attempts to balance the delay required for transmission and retransmission with the delay required for end to end retransmission, such that the overall transfer time is minimized. The response time model factors both the number of packets N to be transmitted and the ....

....transfer time is minimized. The response time model factors both the number of packets N to be transmitted and the round trip time R into the choice of transmission rate, as well as the effects of packet loss, timeouts, and retransmission. The expected response time for the transfer is given by [27]: Response(f) N qf R (1 Gamma log N log p ) where N is the number of packets to be transmitted, R is the round trip time to the destination, p is the packet loss probability when transmitting at rate f , and q = 1 Gamma p. The response time curve is U shaped, with a minimum response ....

[Article contains additional citation context not shown here]

C. Williamson, "Optimizing File Transfer Response Time Using the Loss-Load Curve Congestion Control Mechanism", Proceedings of ACM SIGCOMM, pp. 117-126, September 1993.

....Figures 2.1 ATM Multiplexer : 8 2.2 Multimedia Traffic Characteristics [7] 10 2.3 ATM Traffic Levels [30] 11 2. 4 Two state Markov ATM Traffic Flow Model [31] : 12 2.5 Two state Continuous time Markov Chain Model : 15 4.1 Network Model : 26 5.1 Schedulable Region for Deterministic Scheduling and Deadline Scheduling for the Video and Voice Traffic ....

....strategies can be classified into two groups: reactive and preventive. Reactive approaches rely on the network to provide feedback signals, either explicit or implicit, indicating the presence or absence of congestion. Sources then dynamically adjust the rate of transmission into the network [31]. Preventive approaches rely on reservation of network resources before the sources can access the network, and these sources are required to remain within their allocated resources [18] Reactive strategies are not suitable for an ATM network because of the increased propagation delay bandwidth ....

[Article contains additional citation context not shown here]

C. Williamson, "Optimizing File Transfer Response Time Using the Loss-Load Curve Congestion Control Mechanism," Proceedings of ACM SIGCOMM'93, San Francisco, California, pp 117-126, Sept 1993.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC