J. Nagle, "On packet switches with infinite storage," Communications, IEEE Transactions on [legacy, pre - 1988.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....quality of solution, and or complexity problems. In the following section, we will present the problems of developing a fair bandwidth sharing AQM algorithm and our proposed solution to it. One possible existing solution is to separate flows into different queues, for example, Fair Queueing [6], Weighted Fair Queueing [7] and Generalized Processor Sharing [8] They are defined in terms of a fluid flow model. Packets are assumed to be infinitely divisible and servers are able to serve multiple packets at the same time. A fair queueing scheduler thus provides ideal fairness. For the same ....

J. Nagle, "On Packet Switches with Infinite Storage," IEEE Transactions on Communications, Volume 35, pp 435-438, 1987.

....When long cell bursts from a misbehaving user are queued up in the FIFO, the other regular arrival cells will be delayed or even discarded when the FIFO is full. The round robin discipline is believed to be capable of providing fairer service than the FIFO discipline when the network is congested [7], 81. Their throughput delay per 0733 8716 91 0900 1110501.00 1991 IEEE CHAO: A NOVEL ARCHITECTURE FOR QUEUE MANAGEMENT formance comparison can be found in [9] The roundrobin discipline usually operates by maintaining a separate FIFO queue for each connnection, which could be identified by a ....

.1. B. Nagle, "On packet switches with infinite storages," IEEE Trans. Commun., vol. 35, no. 4 pp. 435-438, Apr. 1987.

....this work is randomized algorithms for fair bandwidth allocation which are simple, practical and robust. As we discuss later, we use randomization to ensure that on the average the bandwidth allocation is fair. Deterministic approaches to packet fair queuing have evolved from the work of Nagle [J. 87] on isolating misbehaving flows. Nagle proposed maintaining a separate FIFO queue for each flow and serving these queues in a round robin fashion. This scheme, 89 although constituting the basis for fair bandwidth allocation, suffered from two major drawbacks: a) Flow with larger packet size ....

....removal and addition of classes. 0 0.2 0.4 0.6 0.8 1 0 5 10 15 20 25 30 35 40 45 50 Figure 5.15: Link sharing with Enhanced Credit PrFQ. In this work we introduced randomized approaches for packet fair queuing. We described algorithms that maintain the isolation concept of fair queuing [J. 87] while handling bandwidth allocation with minimal state per flow. The main contribution of this work is efficient and simple to implement algorithms to handle packet fair queuing. A time complexity of O(1) for the Quanta PrFQ and Credit PrFQ algorithms is achieved using additional data ....

J. Nagle. On packet switches with infinite storage. IEEE Transactions on Communications, 35(4):435--438, April 1987. Also available as RFC 970.

....rate. An application using a transport protocol with flow control will experience the throughput to match the bandwidth specified at the emulator. However, some algorithms on network layer will not work any more like expected: If a real link is saturated, queueing algorithms like fair queueing [13] will try to give every sender its fair share of the resource. If a link is throttled by a common emulator, however, the additional drop tail queue would drain the other queues immediately and prevent the effects of fair queueing (see Fig. 2) This means that the emulation process is not ....

J. Nagle, "On Packet Switches With Infinite Storage," IEEE Transactions on Communications, vol. 35, no. 4, pp. 435-438, April 1987.

....for packet scheduling, especially in the presence of non uniform load. When a distributed queueing system is controlled using a centralised scheduler, an obvious mechanism of granting channel access to mobile stations is polling. One possible strategy for managing the polling is round robin (RR) [11], which polls stations cyclically, regardless of the state of their queues. When RR is used to schedule service from a centralized queue, the scheduler can efficiently bypass stations that have no data to transmit. In contrast, the scheduler of a wireless network must poll a station in order to ....

J. Nagle, "On packet switches with infinite storage," IEEE Trans. Comm., Apr. 1987, vol. COM-35, no. 4. pp. 435-38

....evaluate Faqade in Section 4, both on general SLOs and for array virtualization; we then conclude in Section 5. 2 Related work There is an extensive body of work on networking SLAs [3] and network flow control methods going back to the leaky bucket throttling algorithm [23] and fair queueing [18, 9]. There are several architectures for providing different levels of services to different flows, including the Differentiated Services (DiffServ) architecture [4] and the QoSBox [7] by mapping groups of flows with similar requirements into a few classes of traffic. Unfortunately, most of the ....

J. Nagle. On packet switches with infinite storage. IEEE Transactions on Communications, 35(4):435438, 1987.

....increase the mean delay of packets belonging to flows from other sources. An alternate technique is Packet Based Round Robin (PBRR) in which packets are queued separately, based on the flows to which they belong, and the scheduler transmits one packet from each queue in a round robin fashion [2]. The PBRR scheduler, however, is not fair since flows sending longer packets can use up an unfairly high fraction of the available bandwidth. One of the difficulties in designing a fair scheduling algorithm for interconnection networks of parallel systems is the unique restriction on the ....

.... by the number of packets that are already waiting in the queue and is thus bounded only by the size TABLE 1 Comparison of Fair Scheduling Algorithms Applicable Scheduling Discipline Complexity Fairness Start up Latency Bound to Wormhole Networks GPS [ 16] 0 Packet Based Round Robin [2] O(1) oc First Come First Served O(1) oc Weighted Fair Queueing [14] O(log n) O(n) 2Tn Self Clocked Fait Queueing [17] O(log n) 2m Worst Case Fair Weighted Fair Queueing [21 ] O(log n) 2rr (M m 1)n Deficit Round Robin [22] O(1) g 2rr (3 m 1)n Surplus Round ....

J. Nagle, "On Packet Switches with Infinite Storage," IEEE Trans. Comm., vol. 35, no. 4, Apr. 1987.

....User Datagram Protocol (UDP) that do not back off under congestion. As a result, these congestion unresponsive flows tend to use up bandwidth more aggressively, squeezing out flows that are congestion responsive. This problem of responsive flows versus unresponsive flows was first noted by Nagle [1], who introduced a fair bandwidth sharing scheduling algorithm to alleviate this problem. Subsequently, other researchers also realized the importance of fair bandwidth sharing and how it can greatly improve the performance of end to end congestion control algorithms. Until recently, fair ....

J. Nagle, On packet switches with infinite storage, IEEE Trans. Communications, Vol 35, No. 4, pp 435-438, Apr. 1987.

....level, although a commonly used series 48 of control mechanism exists between that of burst and of call. The RTT, the period of time required for information to be communicated from source to destination and back to source, is a useful timescale to categorise mechanisms such as TCP ( Postel81b, Nagle85, Lottor88, Stevens97, Mathis96] and explicit congestion notification schemes proposed as alternatives to it [Ramakrishnan90, Floyd94, Ramakrishnan99, Key99, Gibbens99] Section 2.3.4 summarises the significant quantity of literature that exists describing control techniques at the call level, ....

J. Nagle. On packet switches with infinite storage. RFC 970, IETF, December

....Research Labs Stanford University ACIRI Abstract Many researchers have argued that the Internet architecture would be more robust and more accommodating of heterogeneity if routers allocated bandwidth fairly. However, most of the mechanisms proposed to accomplish this, such as Fair Queueing [16], 6] and its many variants [2] 23] 15] involve complicated packet scheduling algorithms. These algorithms, while increasingly common in router designs, may not be inexpensively implementable at extremely high speeds; thus, finding more easily implementable variants of such algorithms may be ....

....of our approach is aided by the fact that the vast majority of Internet flows are slow but the fast flows send the bulk of the bits. This allows a small sample of recent history to provide accurate rate estimates of the fast flows. I. INTRODUCTION Since the pioneering observations of Nagle [16], many researchers have argued that the Internet architecture would be more robust (in the face of ill behaved flows) and more accommodating of heterogeneity (by no longer requiring adherence to a single congestion control algorithm) if routers allocated bandwidth fairly. This viewpoint is not ....

[Article contains additional citation context not shown here]

Nagle, J., "On packet switches with infinite storage", Internet Engineering Task Force, RFC-970, December, 1985.

No context found.

J. Nagle, "On packet switches with infinite storage," Communications, IEEE Transactions on [legacy, pre - 1988.

No context found.

John Nagle. On packet switches with infinite storage. RFC 970, Internet Engineering Task Force, December 1985.

No context found.

J. Nagle. On packet switches with infinite storage. IEEE Transacions on Communications, Vol. 35, N 4, pp. 435-438, April 1987.

No context found.

J. Nagle, "On packet switches with infinite storage," IEEE Transactions on Communications, vol. 35, no. 4, pp. 435--438, April 1987, Also available as RFC 970.

No context found.

J.B. Nagle. On packet switches with infinite storage. IEEE Transactions on Communications, 35(4):435--438, April 1987.

No context found.

Nagle, J., "On Packet Switches with Infinite Storage", RFC-970, Ford Aerospace, December 1985.

No context found.

J. Nagle, "On packet switches with infinite storage," in IEEE Transaction on Communications, vol. 35, April 1987.

No context found.

J. Nagle, "On packet switches with infinite storage", IEEE Transactions on Communications, COM-35(4):435--438, April 1987.

No context found.

J. Nagle, "On packet switches with infinite storage", IEEE Transactions on Communications, COM-35(4):435--438, April 1987.

No context found.

Nagle, J., "On Packet Switches with Infinite Storage", DARPA Network Working Group Report RFC-970, Ford Aerospace, December 1985.

No context found.

J. Nagle, "On packet switches with infinite storage," in IEEE Transaction on Communications, vol. 35, April 1987.

No context found.

J Nagle. On packet switches with infinite storage. IEEE Transactions On Communications, 35(4):435--438, April 1987.

No context found.

Nagle, John B., "On Packet Switches With Infinite Storage", RFC 970, FACC Palo Alto, December 1985.

No context found.

J. Nagle, "On packet switches with infinite storage," IEEE Trans. Commun. COM35, 435--438 (1987).

No context found.

Nagle, J., "On packet switches with infinite storage", Internet Engineering Task Force, RFC-970, December, 1985.

First 50 documents  Next 50

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