Probabilistic packet scheduling: Achieving proportional share bandwidth allocation for TCP flows

Ming Zhang; Randolph Wang; Larry Peterson; Arvind Krishnamurthy

doi:10.1109/INFCOM.2002.1019418

Probabilistic packet scheduling: Achieving proportional share bandwidth allocation for TCP flows

Ming Zhang, Randolph Wang, Larry Peterson, Arvind Krishnamurthy

Research output: Contribution to journal › Article › peer-review

15 Scopus citations

Abstract

This paper describes and evaluates a probabilistic packet scheduling (PPS) algorithm for providing different levels of service to TCP flows. With our approach, each router defines a local currency in terms of tickets and assigns tickets to its inputs based on contractual agreements with its upstream routers. A flow is tagged with tickets to represent the relative share of bandwidth it should receive at each link. When multiple flows share the same bottleneck, the bandwidth that each flow obtains is proportional to the relative tickets assigned to that flow. Simulations show that PPS does a better job of proportionally allocating bandwidth than DiffServ and weighted CSFQ. In addition, PPS accomodates flows that cross multiple currency domains.

Original language	English (US)
Article number	55
Pages (from-to)	1650-1659
Number of pages	10
Journal	Proceedings - IEEE INFOCOM
Volume	3
DOIs	https://doi.org/10.1109/INFCOM.2002.1019418
State	Published - 2002
Externally published	Yes

Keywords

Flow control
Quality of service
Scheduling

ASJC Scopus subject areas

General Computer Science
Electrical and Electronic Engineering

Access to Document

10.1109/INFCOM.2002.1019418

Cite this

@article{40732331dbcb48b4b60c58fc3775d3da,

title = "Probabilistic packet scheduling: Achieving proportional share bandwidth allocation for TCP flows",

abstract = "This paper describes and evaluates a probabilistic packet scheduling (PPS) algorithm for providing different levels of service to TCP flows. With our approach, each router defines a local currency in terms of tickets and assigns tickets to its inputs based on contractual agreements with its upstream routers. A flow is tagged with tickets to represent the relative share of bandwidth it should receive at each link. When multiple flows share the same bottleneck, the bandwidth that each flow obtains is proportional to the relative tickets assigned to that flow. Simulations show that PPS does a better job of proportionally allocating bandwidth than DiffServ and weighted CSFQ. In addition, PPS accomodates flows that cross multiple currency domains.",

keywords = "Flow control, Quality of service, Scheduling",

author = "Ming Zhang and Randolph Wang and Larry Peterson and Arvind Krishnamurthy",

year = "2002",

doi = "10.1109/INFCOM.2002.1019418",

language = "English (US)",

volume = "3",

pages = "1650--1659",

journal = "Proceedings - IEEE INFOCOM",

issn = "0743-166X",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Probabilistic packet scheduling

T2 - Achieving proportional share bandwidth allocation for TCP flows

AU - Zhang, Ming

AU - Wang, Randolph

AU - Peterson, Larry

AU - Krishnamurthy, Arvind

PY - 2002

Y1 - 2002

N2 - This paper describes and evaluates a probabilistic packet scheduling (PPS) algorithm for providing different levels of service to TCP flows. With our approach, each router defines a local currency in terms of tickets and assigns tickets to its inputs based on contractual agreements with its upstream routers. A flow is tagged with tickets to represent the relative share of bandwidth it should receive at each link. When multiple flows share the same bottleneck, the bandwidth that each flow obtains is proportional to the relative tickets assigned to that flow. Simulations show that PPS does a better job of proportionally allocating bandwidth than DiffServ and weighted CSFQ. In addition, PPS accomodates flows that cross multiple currency domains.

AB - This paper describes and evaluates a probabilistic packet scheduling (PPS) algorithm for providing different levels of service to TCP flows. With our approach, each router defines a local currency in terms of tickets and assigns tickets to its inputs based on contractual agreements with its upstream routers. A flow is tagged with tickets to represent the relative share of bandwidth it should receive at each link. When multiple flows share the same bottleneck, the bandwidth that each flow obtains is proportional to the relative tickets assigned to that flow. Simulations show that PPS does a better job of proportionally allocating bandwidth than DiffServ and weighted CSFQ. In addition, PPS accomodates flows that cross multiple currency domains.

KW - Flow control

KW - Quality of service

KW - Scheduling

UR - http://www.scopus.com/inward/record.url?scp=0036343748&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036343748&partnerID=8YFLogxK

U2 - 10.1109/INFCOM.2002.1019418

DO - 10.1109/INFCOM.2002.1019418

M3 - Article

AN - SCOPUS:0036343748

SN - 0743-166X

VL - 3

SP - 1650

EP - 1659

JO - Proceedings - IEEE INFOCOM

JF - Proceedings - IEEE INFOCOM

M1 - 55

ER -

Probabilistic packet scheduling: Achieving proportional share bandwidth allocation for TCP flows

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this