R. Guerin and H. Schulzrinne. Network quality of service. In [16], pages 479--503.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....by the user or community. The underlying Grid RM architecture must remain open enough to support multiple concurrent brokering strategies across resources that might be shared by multiple user communities. 7 Other Related Work Numerous researchers have investigated approaches to QoS delivery [22]and resource reservation for networks [12, 14, 39] CPUs [24] and other resources. Proposals for advance reservations typically employ cooperating servers that coordinate advance reservations along an end to end path [39, 14, 12, 23] Techniques have been proposed for representing advance ....

R. Guerin and H. Schulzrinne. Network quality of service. In [16], pages 479--503.

....foreground ows with immediate reservations. In this experiment we used a simple data transfer program, which adapted its transmission rate to the rate assigned to the background class. 7 Related Work The general problem of QoS implementation and management is receiving increased attention (see [10]) Some groups have investigated the use of DS mechanisms (e.g. 23, 19] and its impact on TCP. Sahu et al. 19] in particular, present a quantitative comparison of di erent DS router mechanisms. However, none of the studies has explicitly addressed ows consuming a goodput of 10 MB s or higher ....

Roch Guerin and Henning Schulzrinne. Network quality of service. In [7], pages 479-503.

....and authentication mechanisms provided in the Globus toolkit provide a basis on which to explore these issues. The Akenti system [17] also provides important relevant technology. 7 Related Work The general problem of QoS implementation and management is receiving increased attention (see, e.g. [9]) However, there has been little work on the specific problems addressed in this paper, namely advance reservation and co reservation of heterogeneous collections of resources for end to end QoS and the use of DS mechanisms to support flow types encountered in highend applications. Proposals for ....

Roch Gu'erin and Henning Schulzrinne. Network quality of service. In [7], pages 479--503.

....networks with RSVP signaling. We present experimental results that quantify the cost of local reservation and object creation operations. 2 Related Work The general problem of resource management in networks and wide area computing systems is receiving increased attention (for reviews, see e.g. [1, 10]) However, there has been little work on the specific problems addressed in this paper, namely advance reservation and coallocation of heterogeneous collections of resources for end to end QoS. Here we review briefly some relevant work; space constraints prevent a complete survey. Proposals for ....

Roch Guerin and Henning Schulzrinne. Network quality of service. In [9], pages 479--503.

....Section 4 discusses some of the issues. Finally, Section 5 describes options for billing for Internet telephony. In this brief survey, we do not touch upon the crucial issue of providing quality of service appropriate for telephony services, as these are discussed at length in the literature [1]. It should be noted that we follow convention in using the term Internet telephony, which by its root indicates voice communications. However, unless we are discussing the interoperation with the existing telephone network, this encompasses all media types, as discussed below in Section 2. 2 ....

R. Guerin and H. Schulzrinne, "Network quality of service," in Computational Grids: The Future of High-Performance Distributed Computing (I. Foster and C. Kesselman, eds.), San Francisco, California: Morgan Kaufmann Publishers, 1998.

....translate into a less frequent clicking sound and become less annoying. A di erent example is for video streams. A video frame often consists of several network packets, and losing one packet renders the whole frame useless. In such case bursty losses may actually be preferable to random losses [8]. 5 Conclusion We discussed factors a ecting real time multimediaQoS. The rst is the modeling of network delay and loss. We propose the joint use of the n state extended Gilbert model and inter loss loss distance (ILD) to characterize loss burstiness. This is con rmed by comparing the errors in ....

Roch Gurin and Henning Schulzrinne. Network quality of service. In Ian Foster and C. Kesselman, editors, The Grid: Blueprint for a New Computing Infrastructure. Morgan Kaufmann Publishers, San Francisco, California, 1998.

....hear A s voice clearly. For one way delay, ITU T G.114 [16] recommends 150ms as the upper limit for most applications, 150 to 400ms as potentially intolerable, and above 400 ms as generally unacceptable delay. The one way delay tolerance for video conferencing is in a similar range, 200 to 300ms [14]. Two way metrics are also important, especially for interactive applications such as IP telephony. It is well known that a large RTT ( 600ms) will degrade the interactivity of an application [9] However, if we obtain a good estimate of one way delay in both directions, a simple addition easily ....

.... 3 1 0 1 1 (c) Without FEC, using Prev Opt (d) With FEC, using (5,3) Reed Solomon code and virtualized Prev Opt Table 1: E ect of playout control on nal loss burstiness, CU GMD Sep 1997 trace consists of several network packets, bursty losses may actually be more preferable than random losses [14]. 8 Conclusion We discussed several topics in real time multimedia QoS measurement. The rst is how to obtain more precise measurement data, in particular, how to overcome problems such as clock skew and synchronization. The second topic is the modeling of network delay and loss. The n state ....

Roch Gurin and Henning Schulzrinne. Network quality of service. In Ian Foster and C. Kesselman, editors, The Grid: Blueprint for a New Computing Infrastructure. Morgan Kaufmann Publishers, San Francisco, California, 1998.

....hear A s voice clearly.For one way delay, ITU T G.114 [16] recommends 150ms as the upper limit for most applications, 150 to 400ms as potentially intolerable, and above 400 ms as generally unacceptable delay. The one way delay tolerance for video conferencing is in a similar range, 200 to 300ms [14]. Two way metrics are also important, especially for interactive applications such as IP telephony.Itis well known that a large RTT (# 600ms) will degrade the interactivity of an application [9] However, if we obtain a good estimate of one way delay in both directions, a simple addition easily ....

.... 3 1 0 1 1 (c) Without FEC, using Prev Opt (d) With FEC, using (5,3) Reed Solomon code and virtualized Prev Opt Table 1: E ect of playout control on nal loss burstiness, CU GMD Sep 1997 trace consists of several network packets, bursty losses may actually be more preferable than random losses [14]. 8 Conclusion We discussed several topics in real time multimedia QoS measurement. The rst is how to obtain more precise measurement data, in particular, howtoovercome problems such as clockskew and synchronization. The second topic is the modeling of network delay and loss. The # state ....

Roch Gurin and Henning Schulzrinne. Network quality of service. In Ian Foster and C. Kesselman, editors, ### ##### ######### ### # ### ######### ##############. Morgan Kaufmann Publishers, San Francisco, California, 1998.

No context found.

R. Guerin and H. Schulzrinne. Network quality of service. In [16], pages 479--503.

No context found.

R. Guerin and H. Schulzrinne. Network Quality of Service. In I. Foster and C. Kesselman, editors, The Grid: Blueprint for a Future Computing Infrastructure, pages 479--503. Morgan Kaufmann Publishers, 1998.

No context found.

R. Guerin, and H. Schulzrinne, Network Quality of Service, [20] 479--503.

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