M. Reisslein, K. Ross, and S. Rajagopal, "Guaranteeing statistical QoS to regulated traffic: The multiple node case," in IEEE Decision & Control'98, pp. 531--538, 1998.  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... work in [8] researchers have investigated the statistical multiplexing gain by only assuming that flows are statistically independent, and that traffic from each flow is constrained by a deterministic regulator, e.g. by a leaky bucket [5] 8] 7] 9] 10] 12] 16] 17] 19] 20] [21]. Henceforth, we will refer to traffic which satisfies these assumptions as regulated adversarial traffic. In this paper we attempt to provide new insights into the problem of determining the multiplexing gain of statistically independent, regulated, but otherwise arbitrary traffic flows at a ....

....from multiple flows passing through the same sequence of congested nodes may become correlated, the assumption of statistical independence of flows may not hold in such a setting. Only few results are currently available on end to end QoS guarantees for adversarial regulated traffic [7] 20] [21]. The remaining sections of this paper are structured as follows. In Section II we specify our assumptions on the traffic and define the effective envelopes. In Section III we derive sufficient schedulability conditions for a general class of packet schedulers, which can be used for a ....

M. Reisslein, K. W. Ross, and S. Rajagopal. Guaranteeing statistical QoS to regulated traffic: The single node case. In Proceedings of IEEE INFOCOM'99, pages 1061--1062, New York, March 1999.

.... the seminal work in [8] researchers have investigated the statistical multiplexing gain by only assuming that flows are statistically independent, and that traffic from each flow is constrained by a deterministic regulator, e.g. by a leaky bucket [5] 8] 7] 9] 10] 12] 16] 17] 19] [20], 21] Henceforth, we will refer to traffic which satisfies these assumptions as regulated adversarial traffic. In this paper we attempt to provide new insights into the problem of determining the multiplexing gain of statistically independent, regulated, but otherwise arbitrary traffic flows at ....

....traffic from multiple flows passing through the same sequence of congested nodes may become correlated, the assumption of statistical independence of flows may not hold in such a setting. Only few results are currently available on end to end QoS guarantees for adversarial regulated traffic [7] [20], 21] The remaining sections of this paper are structured as follows. In Section II we specify our assumptions on the traffic and define the effective envelopes. In Section III we derive sufficient schedulability conditions for a general class of packet schedulers, which can be used for a ....

M. Reisslein, K. W. Ross, and S. Rajagopal. Guaranteeing statistical QoS to regulated traffic: The multiple node case. In Proceedings of 37th IEEE Conference on Decision and Control (CDC), pages 531--531, Tampa, December 1998.

....is theoretically attractive, it is not clear whether it can be applied to real life systems since their scheme assumes an infinite buffer at each node. Moreover, the implementation complexity of PGPS must be resolved before it can be used for highspeed networks like ATM [9] Reisslein et al. 21] [22] also showed that statistical delay guarantee for multiplexed real time streams can be achieved in multihop networks by employing cascaded leaky bucket regulators, which adds to the complexity of switches. Effective bandwidth has been investigated in order to provide statistically guaranteed QoS ....

M. Reisslein, K.W. Ross, and S. Rajagopal, "Guaranteeing Statistical (QoS) to Regulated Traffic: The Single Node Case," Proc. IEEE Infocorn '99, pp. 1060-1071, Mar. 1999.

....work is theoretically attractive, it is not clear whether it can be applied to real life systems since their scheme assumes an infinite buffer at each node. Moreover, the implementation complexity of PGPS must be resolved before it can be used for highspeed networks like ATM [9] Reisslein et al. [21], 22] also showed that statistical delay guarantee for multiplexed real time streams can be achieved in multihop networks by employing cascaded leaky bucket regulators, which adds to the complexity of switches. Effective bandwidth has been investigated in order to provide ....

M. Reisslein, K.W. Ross, and S. Rajagopal, "Guaranteeing Statistical (QoS) to Regulated Traffic: The Multiple Node Case," Proc. 37th IEEE Conf. Decision and Control, pp. 531-538, Dec. 1998.

.... that support visual communication and can be introduced with acceptable cost [3, 7] One of the essential questions is whether to design networks with large buffers at the network nodes or rather use small buffers and source shaping when transmitting delay and loss sensitive video traffic (e.g. [8, 9]) Intuitively, using large buffers low packet loss probability can be provided even at high network utilization, but the control of end to end delay and delay variation has to be solved. On the other hand, source shaping with small buffers at the nodes bounds the delays but at the price of ....

....of data in a B frame only affects that particular frame. Source shaping Shapers used at the sources decrease the frame to frame fluctuation of the coded video stream. The shaper we use in this work is a single buffer leaky bucket, as it is proved to be optimal for networks with small buffers [8]. Frames leaving the encoder are stored in the shaper buffer and are transmitted with a given transmission rate, which is adjusted to provide lossless, delay limited shaping. Shaper algorithms for on line video streams are proposed in [6] In this paper, we apply a low complexity solution based on ....

M. Reisslein, K. W. Ross, S. Rajagopal, "Guaranteeing Statistical QoS to Regulated Traffic: The Multiple Node Case," in Proc. IEEE Decision & Control'98, pp. 531-538, 1998.

.... provides enough space to absorb larger bursts and consequently, limits the loss probability while the control of delay and delay variation becomes a complex issue [14] The use of packet scale buffering for transmitting delay and loss sensitive data like coded video streams have been proposed in [6, 15, 16], showing that low packet loss probability and high network utilization can be achieved if the peak rate of the streams is low compared to the link capacities. In [15, 16] packet scale buffering is proposed together with source shaping. It is proved that a single buffer leaky bucket is an optimal ....

....use of packet scale buffering for transmitting delay and loss sensitive data like coded video streams have been proposed in [6, 15, 16] showing that low packet loss probability and high network utilization can be achieved if the peak rate of the streams is low compared to the link capacities. In [15, 16] packet scale buffering is proposed together with source shaping. It is proved that a single buffer leaky bucket is an optimal shaper in this scenario. The shaping of stored variable bit rate video streams is widely analyzed in the literature. Recent solutions are based on network calculus e.g. ....

M. Reisslein, K. W. Ross, S. Rajagopal, "Guaranteeing Statistical QoS to Regulated Traffic: The Single Node Case," in Proc. IEEE Infocom'99, pp. 1061-1072, 1999.

.... provides enough space to absorb larger bursts and consequently, limits the loss probability while the control of delay and delay variation becomes a complex issue [14] The use of packet scale buffering for transmitting delay and loss sensitive data like coded video streams have been proposed in [6, 15, 16], showing that low packet loss probability and high network utilization can be achieved if the peak rate of the streams is low compared to the link capacities. In [15, 16] packet scale buffering is proposed together with source shaping. It is proved that a single buffer leaky bucket is an optimal ....

....use of packet scale buffering for transmitting delay and loss sensitive data like coded video streams have been proposed in [6, 15, 16] showing that low packet loss probability and high network utilization can be achieved if the peak rate of the streams is low compared to the link capacities. In [15, 16] packet scale buffering is proposed together with source shaping. It is proved that a single buffer leaky bucket is an optimal shaper in this scenario. The shaping of stored variable bit rate video streams is widely analyzed in the literature. Recent solutions are based on network calculus e.g. ....

[Article contains additional citation context not shown here]

M. Reisslein, K. W. Ross, S. Rajagopal, "Guaranteeing Statistical QoS to Regulated Traffic: The Multiple Node Case," in Proc. IEEE Decision & Control'98, pp. 531-538, 1998.

....has at least one server in common with the route of the new source. Bufferless multiplexing approach: We saw that QoS provisioning schemes purely based on GPS suffer from scalability problems. Still, the isolation property exhibited by GPS is very attractive. Based on this idea, Reisslein et al. [5, 16] introduced a new approach to define a traffic management scheme. The key idea is to shape the traffic prior to its entrance in the network. The shaping rate is chosen subject to the following constraint: minimizing the rate of the source in the network, or, equivalently, ensuring that the maximum ....

....periods of high activity of the multimedia sessions. Future work should address a possible formulation of DSS in a statistical setting, which would help to admit more QoS sessions. However, the challenge is to keep the scheme scalable. Making use of a statistical multiplexing seems difficult (see [16]) The notion of statistical effective bandwidth, with a scheme remaining deterministic in the core of network, could be a first step in this direction. ....

Reisslein, M., Ross, K., Rajagopal, S.: Guaranteeing statistical QoS to regulated traffic: The multiple node case. In: 37th IEEE Conference on Decision and Control (CDC), Tampa, Florida (1998)

....QoS provisioning scheme, called DSS for Deterministic Shaping Scheme. In Section 4, we review the relevant studies on the GPS policy concerning the delay bounds and the admission control procedures and show their lack of scalability. We also compare our approach with the one of Reisslein et al. [5], which uses the same deterministic shaping but relies on a statistical bufferless multiplexing to achieve a high bandwidth utilization. In Section 5, we evaluate DSS. We first compare DSS with RPPS and show that DSS requires less bandwidth than RPPS. We also evaluate the ability of DSS to ....

....has at least one server in common with the route of the new source. Bufferless multiplexing approach: We saw that QoS provisioning schemes purely based on GPS suffer from scalability problems. Still, the isolation property exhibited by GPS is very attractive. Based on this idea, Reisslein et al. [5, 16] introduced a new approach to define a traffic management scheme. The key idea is to shape the traffic prior to its entrance in the network. The shaping rate is chosen subject to the following constraint: minimizing the rate of the source in the network, or, equivalently, ensuring that the maximum ....

[Article contains additional citation context not shown here]

Reisslein, M., Ross, K., Rajagopal, S.: Guaranteeing statistical QoS to regulated traffic: The single node case. In: Proceedings of IEEE Infocom, New York (1999)

....section. A recent trend on achieving multiplexing gain relies on the assumptions that connections (flows) are statistically independent and smoothed by deterministic regulators at the connections input to the network since statistical characterization of traffic sources is not often reliable [4] [23]. Not surprisingly, this exactly resembles our case. VPN connections are rate controlled based on provisioning policies at the provider edge. In fact, many of the results derived in those will, therefore, be valid in our case too. One interesting result [23] is: by statistically multiplexing rate ....

....sources is not often reliable [4] 23] Not surprisingly, this exactly resembles our case. VPN connections are rate controlled based on provisioning policies at the provider edge. In fact, many of the results derived in those will, therefore, be valid in our case too. One interesting result [23] is: by statistically multiplexing rate controlled (at edge) traffic in the core network the number of accepted connections can be three times higher than that of Generalized Processor Sharing [20] 21] or any other deterministic service discipline [9] The simulation setup that we consider for ....

M. Reisslein, K. W. Ross, and S. Rajagopal. Guaranteeing statistical qos to regulated traffic: The multiple node case. In Proceedings of 37th IEEE Conference on Decision and Control (CDC),Tampa, December 1998.

.... Sharing (GPS) schedulers is developed in [14] End to end statistical QoS guarantees are also provided by the scheme proposed in [25] which employs Traffic Controlled Rate Monotonic Priority Scheduling [24] Our approach was developed independently of [14] 25] and was first presented in [41] [42]. In this article we extend our approach and present it in a comprehensive manner. The GPS based scheme [14] is further refined in [22] Schemes for providing end4o end statistical QoS in a network of Earliest Deadline First (EDF) schedulers are developed in [1] 49] A comparison of the EDF ....

M. Reisslein, K. W. Ross, and S. Rajagopal. Guaranteeing sta- tistical QoS to regulated traffic: The single node case. In Proceedings of IEEE Infocorn '99, pages 1060 1071, New York, NY, March 1999.

.... Sharing (GPS) schedulers is developed in [14] End to end statistical QoS guarantees are also provided by the scheme proposed in [25] which employs Traffic Controlled Rate Monotonic Priority Scheduling [24] Our approach was developed independently of [14] 25] and was first presented in [41], 42] In this article we extend our approach and present it in a comprehensive manner. The GPS based scheme [14] is further refined in [22] Schemes for providing end4o end statistical QoS in a network of Earliest Deadline First (EDF) schedulers are developed in [1] 49] A comparison of the ....

M. Reisslein, K. W. Ross, and S. Rajagopal. Guaranteeing sta- tistical QoS to regulated traffic: The multiple node case. In Proceedings of 37th IEEE Conference on Decision and Control (CDC), pages 531 538, Tampa, FL, December 1998.

....multimedia traffic. In this paper we lay the groundwork for a traffic management architecture that provides end to end statistical QoS guarantees. We focus our attention to a network consisting of a single node in this paper. We extend the traffic management to networks in a subsequent paper [14]. In this paper we view traffic as fluid. The fluid model, which closely approximates a packetized model with small packets, permits us to focus on the central issues and significantly simplifies notation. We suppose that the traffic sent into the node by each connection is regulated by a ....

....(when traffic is packetized) thereby reducing switch cost. A connection s traffic characterization does not change as the traffic passes through the bufferless multiplexer. It is this last feature that is particularly useful when extending the traffic managementscheme to a multihop network [14]. With our scheme the traffic leaving the network node conforms to the same regulator constraints as the traffic entering the node. With shared buffer multiplexers it is difficult (if not impossible) to tightly characterize a connection s traffic once the traffic passes through a shared buffer. ....

[Article contains additional citation context not shown here]

M. Reisslein, K. W. Ross, and S. Rajagopal, "Guaranteeing statistical QoS to regulated traffic: The multiple node case," in Proceedings of 37th IEEE Conference on Decision and Control (CDC), Tampa, FL, Dec. 1998, available at http://www.fokus.gmd.de/usr/reisslein or http://www.eurecom.fr/ross/.

....however, traditional call admission tests typically over provision networking resources. This is because traditional call admission tests usually assume that the connections are adversarial to the extent permitted by the a priori characterizations and transmit worst case traffic patterns [2, 3, 4, 5, 6]. This assumption, however, is often overly conservative, as in most practical circumstances connections do not transmit worst case traffic patterns. As a consequence traditional call admission tests typically over provision networking resources and thus underutilize the network. ....

....enforce per flow QoS; these would require per flow measurements which are not practicable. Therefore, in this paper we focus on measurement 2 based admission rules that provide aggregate QoS. We study the measurement based admission rules within the smoothing bufferless multiplexing framework [5, 6]. The key aspects of the smoothing bufferless multiplexing framework are to (1) pass each connection s traffic through a buffered smoother (peak rate limiter) at the connection s input to the network, and (2) use bufferless statistical multiplexing inside the network. The bufferless multiplexing ....

[Article contains additional citation context not shown here]

M. Reisslein, K. W. Ross, and S. Rajagopal. Guaranteeing statistical QoS to regulated traffic: The multiple node case. In Proceedings of 37th IEEE Conference on Decision and Control (CDC), pages 531--538, Tampa, FL, December 1998.

No context found.

M. Reisslein, K. Ross, and S. Rajagopal, "Guaranteeing statistical QoS to regulated traffic: The multiple node case," in IEEE Decision & Control'98, pp. 531--538, 1998.

No context found.

M. Reisslein, K. Ross, and S. Rajagopal, "Guaranteeing statistical QoS to regulated traffic: The single node case," in IEEE Infocom'99, pp. 1061--1072, 1999.

No context found.

M. Reisslein, K. W. Ross, S. Rajagopal, "Guaranteeing Statistical QoS to Regulated Traffic: The Single Node Case," in Proc. IEEE Infocom'99, pp. 1061-1072, 1999.

No context found.

M. Reisslein, K. W. Ross, S. Rajagopal, "Guaranteeing Statistical QoS to Regulated Traffic: The Multiple Node Case," in Proc. IEEE Decision & Control'98, pp. 531-538, 1998.

No context found.

M. Reisslein, K. W. Ross, and S. Rajagopal. Guaranteeing Statistical QoS to Regulated Traffic: The Single Node Case. In INFOCOM (3), pages 1061--1072, 1999.

No context found.

M. Reisslein, K. Ross, and S. Rajagopal. Guaranteeing statistical QoS to regulated traffic: The multiple node case. In Proceedings of IEEE Conference on Decision and Control, pages 531--538, Tampa, FL, December 1998.

No context found.

M. Reisslein, K. Ross, and S. Rajagopal, "Guaranteeing statistical QoS to regulated traffic: The single node case," in IEEE Infocom'99, pp. 1061--1072, 1999.

No context found.

M. Reisslein, K. W. Ross, S. Rajagopal, "Guaranteeing Statistical QoS to Regulated Traffic: The Single Node Case," in Proc. IEEE Infocom'99, pp. 1061-1072, 1999.

No context found.

M. Reisslein, K. W. Ross, S. Rajagopal, "Guaranteeing Statistical QoS to Regulated Traffic: The Multiple Node Case," in Proc. IEEE Decision & Control'98, pp. 531-538, 1998.

No context found.

M. Reisslein, K. Ross, and S. Rajagopal, "Guaranteeing statistical QoS to regulated traffic: The multiple node case," in IEEE Decision & Control'98, pp. 531--538, 1998.

No context found.

M. Reisslein, K. Ross, and S. Rajagopal, "Guaranteeing statistical QoS to regulated traffic: The single node case," in IEEE Infocom'99, pp. 1061--1072, 1999.

First 50 documents

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