P. Ferguson and G. Huston. Quality of Service: Delivering QoS on the Internet and in Corporate Networks. Wiley & Sons, New York, USA, Jan. 1998.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....where prices fluctuate rapidly. Keywords: computer networks, bandwidth markets, QoS, combinatorial trading, option pricing 1 Introduction Computer networks are now being used to transfer live video and other data that need reliable service from the network, in terms of latency, jitter, and loss [1, 2, 3]. Bandwidth or capacity markets can help to provide the necessary end user quality of service guarantees [4, 5, 6, 7] Capacity trading enables users to reserve capacity in congested networks, and thereby get a guaranteed throughput, even when the network is congested. The networking QoS routing ....

....If the admittance cost is less than the budget, the request is accepted, otherwise it is rejected. The request type determines how the admittance cost is computed. There are three types of requests, spot requests, null broker requests, and derivative broker requests. In a simulation, q [0, 1] of the requests are broker requests, and (1 q) of the requests are spot requests. Derivative brokers and null brokers are not mixed, thus a simulation contains either only spot and derivative broker requests, or only spot and null broker requests. Spot requests are traded immediately, while ....

P. Ferguson and G. Huston. Quality of Service: Delivering QoS on the Internet and in Corporate Networks. John Wiley and Sons, January 1998.

....and analytical isolation bounds for a variety of policies. I. INTRODUCTION There is an increasing demand from the users for having some degree of Quality of Service (QoS) in the new packetswitched networks. This has generated a lot of effort to devise techniques, in order to provide such a QoS [10, 9]. One of the aspects to be controlled in a packet switched network in order to guarantee QoS is the congestion at the switches. Congestion can increase the end to end delay of a packet and can even force the switch to drop packets. One of the means to reduce congestion is to use a good ....

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks. New York: John Wiley & Sons, Inc., 1998.

....findings, and conclusions or recommendations expressed in this paper are those of the authors and do not necessarily reflect the views of the National Science Foundation. developed. Several books have been written on the subject, some focus more on architectural and other practical issues [3], 4] while others on theoretical aspects of QoS provisioning [5] 6] To provide a more focused overview, in this paper we survey a number of recent advances in Internet QoS provisioning, with emphasis on theoretical developments. The objective is two fold: 1) to provide the reader with the ....

P. Ferguson and G. Houston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, John Wiley & Sons, 1998.

....usually pay a at rate. A number of companies have prepared products for QoS (quality of service) i.e. for controlling an Internet like network in which packets are treated di erentially, but the need for such products is not yet proven. A survey of the di erent proposals for QoS may be found in [1]. Since the Internet is a loose organization that is the product of cooperation between very diverse bodies, billing for such services would also be a major problem. Many networking experts have therefore claimed that the current fat dumb pipe model is best and argued that it will prevail due to ....

P. Ferguson and G. Huston. Quality of Service: Delivering QoS on the Internet and in Corporate Networks. Wiley, 1998. 26

....to either inefficient network operation, or to inadequate performance for some applications and users. Because of this issue, there is a wide consensus that the current Internet service model should be replaced with an architecture that can offer different QoS to different traffic types or users [41]. The desired characteristics or objectives of this new service model, however, are not yet clear. The following two sections describe the two major architectures that have been developed within the Internet Engineering Task Force (IETF) 49] for addressing this issue: the Integrated Services ....

....major backbone providers connecting smaller regional networks, which then connect even smaller local networks, and so on. In such a hierarchical structure, establishing the required multilateral agreements in order to enable end to end services has been proven to be quite difficult in practice [41]. Third, IntServ is based on per flow resource reservations, meaning that a certain amount of forwarding resources (bandwidth and buffers) has to be reserved for a flow before the session starts and for the duration of the session. In order for resource reservations to be accomplished in a ....

P. Ferguson and G.Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, John Wiley and Sons, January 1998.

....by the connection at the set up phase. This number might be equal to the average burst size or might be larger than the average burst size (This is determined by the policy and cost. It is expected that the real time VBR traffic is shaped by using either a leaky bucket or a token bucket scheme [14] to constrain the burstiness to within certain limits . In each burst, a field is included to indicate the number of slots required for the next burst (Figure 7) If the aggregate number of slots requested by all the bursting connections, in a particular frame, is less than or equal to the number ....

P.Ferguson and G.Hutson, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, John Wiley

....especially in the areas of packet scheduling and queue management algorithms, have been proposed. Elegant theories such as network calculus and effective bandwidths have also been developed. Several books have been written on the subject, some focus more on architectural and other practical issues [3], 4] while others on theoretical aspects of QoS provisioning [5] 6] To provide a more focused overview, The authors affiliations and contact information are: Victor Firoiu, Nortel Networks, 600 Technology Park Drive, Bilerica, MA 01821 USA, vfiroiu nortelnetworks.com; Jean Yves Le Boudec, ....

P. Ferguson and G. Houston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, John Wiley & Sons, 1998.

....alternative to complex QoS structures and dynamic resource management schemes that are good only for transient congestion. On a global scale, overprovisioning is considered an economically prohibitive luxury since bandwidth is very unlikely to become infinite and cheap in the near term [43]. Even if prices drop, there will always be bottlenecks where resource allocation will have to be regulated and congestion controlled to provide the appropriate levels of service to everyone. Indeed, in the past, similar forecasts based on expectations for increased availability in resources and ....

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, 1998.

....Thus di#erent customers can be provided di#erentiated services in a scalable manner. Therefore these priority schemes are being considered in the Integrated Services as well as Di#erentiated Services Architectures in the Internet (see Braden et al. 4] Blake et al. 3] and Ferguson and Hudson [6]) In this paper we consider a queue with a finite bu#er and two priorities of tra#c. In contrast to other studies, by a finite bu#er of size s we mean that at most s amount of work (e.g. bits) can be stored in the queue at any time. Thus, if a packet of size x arrives but the 1 space available ....

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, N.Y., 1998.

....is, if a class H request arrives when a class L request is being served, then the class H request gets served immediately. The suspended service for the class L request is resumed later once there is no class H request to be served. The preemptiveway is not the only manner to realize priorities [11], but this assumption will not impair the generality of our conclusion as well as will simplify the analysis. With total arrival rate and service rate , the user s requests can be submitted to class H with a probability of ff H (s) or submitted to class L with ff L (s) or discarded with 1 ; ff H ....

P.Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley,New York, 1998.

....the perceived performance, and there appears to be a wide consensus that new applications, especially real time ones such as packet telephony, will require changing how the Internet operates. Various QoS (quality of service) techniques are being developed. For a general survey and references, see [12]. They will provide differentiated service levels. Many of these schemes are complicated, and involve substantial costs in both development and operations. Furthermore, since the basic problem is that of allocating a limited resource, any solutions will surely have to involve a pricing mechanism. ....

.... priority packets to block completely lower priority ones violates the fairness criterion that appears to be important to consumers (see Section 5 for further discussion of this topic) A better approach might be to use weights in routing decisions, such as in the weighted round robin technique [12]. One could also use different approaches in different parts of the network. One can even mix these approaches on the same link. In general, assignments of capacities and prices to the channels in PMP should stay constant for extended periods. This would fit consumer preferences for simplicity ....

Ferguson, P. and Huston, G. Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, 1998.

....of TCP acknowledgements. STARS makes the two reservations for a bidirectional flow simultaneously. Finally, we are aware that network QoS may be provided by any of several means. STARS can take advantage of whatever service differentiation mechanism RSVP [2] DiffServ [1] MPLS [3] ATM QoS [6], etc. is available in the network, because STARS is a framework on top of these mechanisms. II.B. Domains and users The domain whence a reservation request comes is the originating domain; by extension, the domain of the other endpoint in the request is the destination domain. One of STARS ....

P. Ferguson and G. Huston. Quality of Service: Delivering QoS on the Internet and in Corporate Networks. John Wiley & Sons, Inc., 1998.

....pay a at rate. A number of companies have prepared products for QoS (quality of service) i.e. for controlling an Internet like network in which packets are treated di erentially, but the need for such products is not yet proven. A 2 survey of the di erent proposals for QoS may be found in [1]. Since the Internet is a loose organization that is the product of cooperation between very diverse bodies, billing for such services would also be a major problem. Many networking experts have therefore claimed that the current fat dumb pipe model is best and argued that it will prevail due to ....

P. Ferguson and G. Huston. Quality of Service: Delivering QoS on the Internet and in Corporate Networks. Wiley, 1998.

....every two years, while total costs increase much more slowly, so that the price per unit of bandwidth decreases rapidly, it might make sense to provide uniformly high QoS for everybody and avoid the complexities of the schemes that are being considered. Existing work on QoS, surveyed in the book [FergusonH], does not contain any projections of the degree to which the different proposals for providing QoS will lower network utilization. The relation between utilization of network capacity and perceived quality of service is 3 Table 1: Intel and its microprocessors. For each year we list the ....

....switches are fast enough already that if congestion does not cause buffers to fill up, the quality is sufficient for all anticipated demands. In this work, we will assume as a first approximation that improved QoS is associated directly with low utilization levels. Although schemes like those in [FergusonH] can increase the efficiency of the network, whether it has just a single best effort service, or several classes of service, it is hard to incorporate them into an economic model until more is known about their performance. To explore potential futures for QoS on the Internet with and without ....

[Article contains additional citation context not shown here]

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, 1998.

....Saturday, and Sunday each having its own particular load graph. This is the same behavior that has been observed on the switched voice network. There are many inefficiencies in data networks that are not being exploited. Most attention is currently devoted to Quality of Service (QoS) measures [4], but without providing quantitative estimates of how much such measures will save, or will improve the quality of transmission. However, there are many other steps that can be taken to provide a better Internet, steps whose benefits can often be quantified much more easily and reliably. For ....

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, 1998.

....the Xing [9] and Netshow of the Microsoft [6] These technologies provide best effort movie data delivery over TCP IP based on proprietary movie format. The current implementation of the TCP IP inherently does not provide any guaranteed Quality of Service (QoS) for delivering such video streams [2, 8]. Since each method provides the client with only one data stream from one server, it often suffers from poor quality of pictures in the case of network link congestion. It is therefore highly demanded to develop methods of efficiently delivering high quality movie stream that is less ....

....traditional best effort single streaming methods. The initial experimental results warrant further research in this direction to provide high quality video delivery including: studies on the effect of skipping GOPs in the existence of network link congestions; multithreaded GOP delivery over RSVP [2]. 6. ACKNOWLEDGMENTS The research presented in this paper has been in part supported by DOE (DE FG02 97ER25339) and DOD (N62306 99 DB004) 7. ....

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley Computer Publishing, 1998.

....terminals. CEC Deliverable No: AC085 NMP R T DR P 032 b1 Title of Deliverable: 1D6 IP over wireless ATM AC085 NMP DS P 239 1.0 14 5. QUALITY OF SERVICE MANAGEMENT 5. 1 Internet QoS Internet QoS can be expressed as the combination of network imposed delay, jitter, bandwidth and reliability [Fer98]. Delay is the elapsed time for a packet to traverse from the source to the destination. Jitter is the perceived variation in end to end delay. Bandwidth is the maximal data transfer rate that can be sustained between source and destination. Finally, reliable transmission delivers all the packets ....

....of the Integrated Services Architecture. It allows users to communicate their QoS requirements to routers on the data path. Thus RSVP requires (soft) connection state maintenance in each router along the data path. The disadvantages of RSVP relate to the scalability of Internet backbone operation [Fer98]. First of all, RSVP state maintenance in routers consumes a lot of memory. Secondly, periodic state refresh messages increase the Internet traffic load. Recently, it has been recognised that the integrated services architecture signalling protocol, RSVP, can interoperate with other QoS ....

P. Ferguson, G. Huston, "Quality of Service, Delivering QoS on the Internet and in Corporate Networks", Wiley Computer Publishing, ISBN 0-471-24358-2, January 1998.

....bandwidth brokers, the merging of routing and QoS, traffic engineering, constraint based routing and multiprotocol label switching (MPLS) as well as end host support for QoS. We identify some important design principles and open issues for Internet QoS. 1 Introduction Quality of Service (QoS) [15, 17] has been one of the principal topics of research and development in packet networks for many years. This paper presents an overview of Internet QoS. The paper is structured as follows: Section 2 discusses motivation and special considerations for adding QoS to the Internet. Section 3 defines ....

P. Ferguson and G. Huston. Quality of Service: Delivering QoS in the Internet and the Corporate Network. Wiley Computer Books, New York, NY, 1998.

....Quality of Service (QoS) on Internet2, it is likely that efforts from the measurement working group will be devoted into QoS measurement. QoS is probably the most elusive, confounding, and confusing term used in networking community today. There is an interesting discussion on this subject in [9]. The confusion on QoS arises primarily because QoS means different things to different people. Therefore, it is essential that we must define what QoS means to us before we make any attempt for measurement. We believe that the QoS metrics and measurement methodologies in Internet2 should be ....

P. Ferguson and G. Huston, Quality of Service -- Delivering QoS on the Internet and in Corporate Networks, Chapter 1, John Wiley & Sons, Inc., 1998.

....the perceived performance, and there appears to be a wide consensus that new applications, especially real time ones such as packet telephony, will require changing how the Internet operates. Various QoS (quality of service) techniques are being developed. For a general survey and references, see [FergusonH]. They will provide differentiated service levels. Many of these schemes are complicated, and involve substantial costs in both development and operations. Furthermore, since the basic problem is that of allocating a limited resource, any solutions will surely have to involve a pricing mechanism. ....

.... priority packets to block completely lower priority ones violates the fairness criterion that appears to be important to consumers (see Section 5 for further discussion of this topic) A better approach might be to use weights in routing decisions, such as in the weighted round robin technique [FergusonH]. One could also use different approaches in different parts of the network. One can even mix these approaches on the same link. In general, assignments of capacities and prices to the channels in PMP should stay constant for extended periods. This would fit consumer preferences for simplicity ....

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, 1998.

....are small, and what matters is how quickly the entire page is delivered. Therefore ATM is irrelevant from users perspective, and finds its greatest applications in core networks, where aggregate traffic flows do resemble the traffic conditions for which ATM was designed. Most QoS measures (see [FergusonH] for a survey) are also of doubtful utility in the current environment. In an environment such as that of figures 2, 4, and 5, it is intuitively appealing to create a special lane for high priority traffic. In the environment of Fig. 3, though, which is much more representative of the universe of ....

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, 1998.

....only a fixed monthly fee for the access link to the network. However, there are signs of an imminent change. There is extensive work on provision of Quality of Service (QoS) with some transmissions getting preferential treatment. For a survey of this area and references, see the recent book [FergusonH]. Differential service will likely require more complicated pricing schemes, which will introduce yet more complexity. The motivation behind the work on QoS is the expectation of continued or worsening congestion. As Ferguson and Huston say (p. 9 of [FergusonH] it sometimes is preferable to ....

....and references, see the recent book [FergusonH] Differential service will likely require more complicated pricing schemes, which will introduce yet more complexity. The motivation behind the work on QoS is the expectation of continued or worsening congestion. As Ferguson and Huston say (p. 9 of [FergusonH]) it sometimes is preferable to simply throw bandwidth at congestion problems. On a global scale, however, overengineering is considered an economically prohibitive luxury. Within a well defined scope of deployment, overengineering can be a cost effective alternative to QoS structures. The ....

[Article contains additional citation context not shown here]

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, 1998.

.... 650 91 486 DX (50 MHz) 41 644 4779 819 92 DX2 (66 MHz) 54 600 5844 1067 93 Pentium (66 MHz) 112 898 8782 2295 94 Pentium (100 MHz) 166 935 11521 2266 95 Pentium Pro (200 MHz) 400 1325 16202 3566 96 20847 5157 97 Pentium II (300 MHz) 600 735 25070 8945 Existing work on QoS, surveyed in the book [FergusonH], does not contain any projections of the degree to which the different proposals for providing QoS will lower network utilization. The relation between utilization of network capacity and perceived quality of service is a complex one. It is possible to have a lightly utilized network that ....

....switches are fast enough already that if congestion does not cause buffers to fill up, the quality is sufficient for all anticipated demands. In this work, we will assume as a first approximation that improved QoS is associated directly with low utilization levels. Although schemes like those in [FergusonH] can increase the efficiency of the network, whether it has just a single best effort service, or several classes of service, it is hard to incorporate them into an economic model until more is known about their performance. To explore potential futures for QoS on the Internet with and without ....

[Article contains additional citation context not shown here]

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, 1998.

....every two years, while total costs increase much more slowly, so that the price per unit of bandwidth decreases rapidly, it might make sense to provide uniformly high QoS for everybody and avoid the complexities of the schemes that are being considered. Existing work on QoS, surveyed in the book [4], does not contain any projections of the degree to which the different proposals for providing QoS will lower network utilization. The relation between utilization of network capacity and perceived quality of service is a complex one. It is possible to have a lightly utilized network that ....

....MHz) 54 600 5844 1067 93 Pentium (66 MHz) 112 898 8782 2295 94 Pentium (100 MHz) 166 935 11521 2266 95 Pentium Pro (200 MHz) 400 1325 16202 3566 96 20847 5157 97 Pentium II (300 MHz) 600 735 25070 8945 improved QoS is associated directly with low utilization levels. Although schemes like those in [4] can increase the efficiency of the network, whether it has just a single best effort service, or several classes of service, it is hard to incorporate them into an economic model until more is known about their performance. To explore potential futures for QoS on the Internet with and without ....

[Article contains additional citation context not shown here]

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, 1998.

....appears to be a wide consensus that new applications, especially real time ones such as packet telephony, will require changing how the Internet operates. Various QoS (quality of service) techniques are being developed and are beginning to be deployed. For a general survey and references, see [4]. They will provide differentiated service levels. Many of these schemes are complicated, and involve substantial costs in both development and operations. Furthermore, since the basic problem is that of allocating a limited resource, it is widely accepted that all solutions will have to involve ....

....aggregating all the traffic, and also fits consumer preferences for simple schemes. The basic version of PMP mentioned in the Introduction assigns to each channel a fixed fraction of the capacity of the entire network. One can also use weighted priorities, as in the weighted round robin technique [4]. The advantage of the priority approach is that the full gain from aggregating all traffic on one network would be obtained. However, allowing high priority packets to block completely lower priority ones violates the fairness criterion that appears to be important to consumers. In general, ....

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, 1998.

....in the Di erentiated Services (diffserv) Architecture [31, 32, 33] individual user level ows are assigned to only a small number of prede ned classes. Packets in each class are served by a class based scheduling policy, such as class based static priority, or class based weighted fair queuing [4, 5]. As a result, nodes are aware only of aggregations of ows. Each edge router is supposed to aggregate the individual ows into a small number of such aggregate ows. While the bene ts of ow aggregation in terms of reduced overhead for packet scheduling are clear, the question of how to take ....

P. Ferguson and G. Houston, \Quality of services - Delivering QoS on the Internet and in Corporate Networks," Wiler Computer Publishing, 1998

....in the Differentiated Services (diffserv) Architecture [18] individual user level flows are assigned to only a small number of predefined classes. Packets in each class are served by a classbased scheduling policy, such as class based static priority, or class based weighted fair queuing [4]. While the benefits of flow aggregation in terms of reduced overhead for packet scheduling are clear, the question of how to take advantage of flow aggregation for scaling of admission control remains to be answered. In this paper, we address this question in the content of a differentiated ....

P. Ferguson and G. Houston, "Quality of services - Delivering QoS on the Internet and in Corporate Networks, " Wiler Computer Publishing, 1998.

No context found.

P. Ferguson and G. Huston. Quality of Service: Delivering QoS on the Internet and in Corporate Networks. Wiley & Sons, New York, USA, Jan. 1998.

No context found.

Ferguson, P. and G. Huston (1998). Quality of Service: Delivering QoS on the Internet and in Corporate Networks. Wiley.

No context found.

P.Ferguson, G.Huston, Quality of Service: delivering QoS on the Internet and in corporate networks, John Wiley & Sons, 1998.

No context found.

P. Ferguson and G. Huston. Quality of Service: Delivering QoS on the Internet and in Corporate Networks. John Wiley and Sons, January 1998.

No context found.

P. Ferguson, G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, John Wiley & Sons, Inc., 1998.

No context found.

P. Ferguson and G. Huston. Quality of Service: Delivering QoS on the Internet and in Corporate Networks. Wiley & Sons, New York, USA, Jan. 1998.

No context found.

P. Ferguson and G. Huston. Quality of Service: Delivering QoS on the internet and in corporate networks. John Wiley & Sons Inc., 1998.

No context found.

P. Ferguson and G. Houston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, John Wiley & Sons, 1998.

No context found.

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, 1998.

No context found.

P. Ferguson and G. Huston, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, Wiley, 1998.

No context found.

Fergusson, P., Huston, G., " Quality of Service: delivering QoS on the Internet and in Corporate Networks", Wiley Computer Publishing, 1998.

No context found.

Ferguson, Paul and Huston, Geoff, Quality of Service: Delivering QoS on the Internet and in Corporate Networks, New York: Wiley, 1998.

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