Analysis and simulation of a fair queueing algorithm

DemersA.,; KeshavS.,; Shenker, Scott

doi:10.1145/75247.75248

Cited by 1,550 publications

(736 citation statements)

References 11 publications

Supporting

Mentioning

726

Contrasting

Unclassified

Order By: Relevance

“…For C=3, multiple scheduling scenarios are possible. For example, if we consider Residual Voice, Video Telephony and Interactive Games, then we haver=[300,400,500] ≡ [3,4,5].…”

Section: Numerical Resultsmentioning

confidence: 99%

“…These packetized implementations of GPS are known as Packetized GPS (or PGPS) [5] or Weighted-Fair Queueing [4]. Under the assumption of continuously-backlogged users and constant packet (slot) size, it can be shown that PGPS schemes are equivalent to the WRR-sp K scheduler.…”

Section: Performance Comparison With Gps and Pgps Schedulersmentioning

confidence: 99%

“…Since sending packets in a bit-by-bit fashion is unrealistic, packet-by-packet transmission schemes have been proposed to emulate the BR scheme, most of which are variants of the Weighted-Fair Queueing (WFQ) discipline [4]. This was extended in [5] with the Generalized Processor Sharing (GPS) discipline (and corresponding Packetized GPS (PGPS) schemes), where each user i is associated with its fair share, x i , where x i = x j for users i, j.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Design and analysis of a class-aware recursive loop scheduler for class-based scheduling

Rom

Sidi

Tan

2006

Performance Evaluation

View full text Add to dashboard Cite

In this paper, we consider the problem of devising a loop scheduler that allocates slots to users according to their relative weights as smoothly as possible. Instead of the existing notion of smoothness based on balancedness, we propose a variance-based metric which is more intuitive and easier to compute.We propose a recursive loop scheduler for a class-based scheduling scenario based on an optimal Weighted Round Robin scheduler. We show that it achieves very good allocation smoothness with almost no degradation in intra-class fairness. In addition, we also demonstrate the equivalence between our proposed metric and the balancedness-based metric.

show abstract

“…For C=3, multiple scheduling scenarios are possible. For example, if we consider Residual Voice, Video Telephony and Interactive Games, then we haver=[300,400,500] ≡ [3,4,5].…”

Section: Numerical Resultsmentioning

confidence: 99%

Section: Performance Comparison With Gps and Pgps Schedulersmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Design and analysis of a class-aware recursive loop scheduler for class-based scheduling

Rom

Sidi

Tan

2006

Performance Evaluation

View full text Add to dashboard Cite

show abstract

“…3 We have utilized a simple fair-queuing scheme in developing our solution. To appreciate the novelty of our work in the presence of other literature on fair-queuing [22 23], consider that any fair-queuing scheme requires 3 policies [20] for choosing: 1) Which packet (queue) will be transmitted? For this, we use simple round-robin.…”

Section: Background and Previous Workmentioning

confidence: 99%

Improving the Performance of TCP in the Presence of Interacting UDP Flows in Ad Hoc Networks

Gupta

Krishnamurthy

Faloutsos

2004

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract.In this paper, we study how fairness affects the performance of TCP over ad hoc networks with IEEE 802.11 at the MAC layer. The problem addressed is that the throughput of TCP flows degrades severely in the presence of heavily loaded UDP flows. Our contribution is twofold. First, we identify the factors affecting the TCP throughput by providing a micro-analysis of the performance at a level of detail that is not seen in previous studies. The intuition obtained from the first part leads us to our second contribution. We propose and study the use of per flow fairness through a mechanism we call backpressure. Backpressure improves the performance of TCP flows in the presence of heavy UDP flows. In fact, in some cases, this increased TCP throughput does not affect the throughput of UDP flows. We find that backpressure can increase the TCP throughput by as much as 95%. An advantage of backpressure is that it does not require any changes to the existing TCP or IEEE 802.11 protocols.

show abstract

“…A theoretically ideal and fair way to share the server capacity over different flows is Generalised Processor Sharing (GPS) [41,51], but this mechanism is difficult to apply in packet-based networks, so adaptations for packet scheduling are needed. For instance in the framework of ATM [3], weighted-round-robin (WRR) and weighted-fair-queueing (WFQ) [13] were proposed to achieve GPS-like weighted throughput and fairness [33]. For these mechanisms, there are separate queues for each type of traffic and the server 'visits' each queue in a cyclic, weighted and/or timed manner [42].…”

Section: Introductionmentioning

confidence: 99%

Transform-Domain Analysis of Packet Delay in Network Nodes with QoS-Aware Scheduling

Vuyst

Wittevrongel

Bruneel

2011

Network Performance Engineering

View full text Add to dashboard Cite

Abstract. In order to differentiate the perceived QoS between traffic classes in heterogeneous packet networks, equipment discriminates incoming packets based on their class, particularly in the way queued packets are scheduled for further transmission. We review a common stochastic modelling framework in which scheduling mechanisms can be evaluated, especially with regard to the resulting per-class delay distribution. For this, a discrete-time single-server queue is considered with two classes of packet arrivals, either delay-sensitive (1) or delay-tolerant (2). The steady-state analysis relies on the use of well-chosen supplementary variables and is mainly done in the transform domain. Secondly, we propose and analyse a new type of scheduling mechanism that allows precise control over the amount of delay differentiation between the classes. The idea is to introduce N reserved places in the queue, intended for future arrivals of class 1.

show abstract

Analysis and simulation of a fair queueing algorithm

Cited by 1,550 publications

References 11 publications

Design and analysis of a class-aware recursive loop scheduler for class-based scheduling

Design and analysis of a class-aware recursive loop scheduler for class-based scheduling

Improving the Performance of TCP in the Presence of Interacting UDP Flows in Ad Hoc Networks

Transform-Domain Analysis of Packet Delay in Network Nodes with QoS-Aware Scheduling

Contact Info

Product

Resources

About