A new scheduling mechanism to provide relative differentiation for real-time IP traffic

Bodamer, Stefan

doi:10.1109/glocom.2000.892096

Cited by 11 publications

(15 citation statements)

References 19 publications

(19 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example the so called Olympic Service model introduced in [HBWW99] could be implemented by using some variant of WFQ. However, it is stated in [Bod00] and [DSR02] that fair queueing schedulers are sensitive to variations in the class load distribution. If static class weights are used, a priority inversion between classes may easily occur when the load to some trac class increases.…”

Section: Adaptive Wfqmentioning

confidence: 99%

“…Further, it is also possible to assign each class a deadline and dierentiate with the probability that this deadline is violated. For this kind of deadline violation probability dierentiation the WEDD algorithm proposed in [Bod00] would be suitable.…”

Section: Selecting a Packet Scheduler For Delay Dierentiationmentioning

confidence: 99%

“…This scheduler uses many parameters that have to be set correctly in order to receive the desired result. Originally the algorithm was suggested to be used only for real time trac [Bod00]. In this case it could probably be possible to set the deadlines, weights etc.…”

Section: Dierentiation With Hpd and Delay Boundmentioning

confidence: 99%

“…In [Bod00], a new scheduler WEDD is proposed. WEDD is an extension of the classical EDD scheduler in the sense that it provides dierentiation both in terms of delay and deadline violation probabilities.…”

Section: Adaptive Wfqmentioning

confidence: 99%

“…In [Bod00], the model is used in dierentiating the deadline violation probabilities of trac classes. This new model corresponds to the PDD model except that the mean delaysd i are replaced with deadline violation probabilities p i such that…”

Section: Adaptive Wfqmentioning

confidence: 99%

See 4 more Smart Citations

Scheduling and Quality Differentiation in Differentiated Services

Antila¹,

Luoma²

2003

Interactive Multimedia on Next Generation Networks

View full text Add to dashboard Cite

During the last decade the Internet has developed into a public multi service network that should be able to support heterogeneous applications and customers with diverse requirements. For this reason, quality of service (QoS) provisioning in the Internet has gained increasing attention.General service architectures have been proposed in the literature for providing QoS. At the moment, the Dierentiated Services (DiServ) architecture seems to be the most promising solution due to its scalability. In DiServ, common resources are allocated among service classes. The mechanism that is primarily responsible for the allocation is packet scheduling.The DiServ architecture sets certain limits to the services that can be oered to customers. However, a lot of freedom is still left for the operators. For example, the operators can choose what kind of dierentiation models they will use and how these models will be implemented.This thesis studies issues related to quality dierentiation and scheduling in Dierentiated Services networks. The rst part of the thesis provides a literature survey about QoS guarantees, architectures and mechanisms, with the focus on scheduling disciplines.In the second part of the thesis possible service approaches and dierentiation models for an operator are examined and schedulers for implementing these models are proposed based on a qualitative analysis. Some of the proposed dierentiation models and schedulers are evaluated with simulations. The simulations study both at aggregate and session/ow level what kind of service the schedulers provide and how well they are able to support the dierentiation model that they are designed for. In addition, the simulations give insight into the eect of trac mapping on the dierentiation. Palvelun laadun mahdollistamiseksi on ehdotettu erilaisia palveluarkkitehtuureja. Täl-lä hetkellä eriytettyjen palveluiden (DiServ) arkkitehtuuri vaikuttaa lupaavimmalta ratkaisulta skaalautuvuutensa ansiosta. DiServ-arkkitehtuurissa yhteiset resurssit jaetaan palveluluokkien kesken. Mekanismi, joka pääasiassa vastaa resurssien allokoinnista, on pakettien skeduloija.DiServ arkkitehtuuri asettaa tietyt rajoitukset asiakkaille tarjottaville palveluille. Operaattoreille jää kuitenkin runsaasti vapausasteita. Operaattorit voivat esimerkiksi päät-tää, millaisia eriyttämismalleja käytetään ja miten mallit toteutetaan.Tässä työssä tutkitaan laadun eriyttämistä ja skedulointia DiServ-verkoissa. Työn ensimmäinen osa on kirjallisuustutkimus palvelunlaatutakuista, arkkitehtuureista sekä mekanismeista pääpainon ollessa skeduloinnissa.Työn toisessa osassa analysoidaan kvalitatiivisella tasolla operaattorin mahdollisia palvelutapoja ja eriyttämismalleja sekä ehdotetaan erilaisia skedulointimenetelmiä mallien toteuttamiseksi. Malleja ja skedulereita arvioidaan myös simulaatioiden avulla. Simulaatioissa tutkitaan sekä aggregaatti-että sessio/vuotasolla, millaista palvelua eri skedulerit pystyvät tarjoamaan ja miten hyvin ne tukevat valittua eriyttämismallia. Lisäksi tarkastellaan liikenteen ...

show abstract

Section: Adaptive Wfqmentioning

confidence: 99%

Section: Selecting a Packet Scheduler For Delay Dierentiationmentioning

confidence: 99%

Section: Dierentiation With Hpd and Delay Boundmentioning

confidence: 99%

Section: Adaptive Wfqmentioning

confidence: 99%

Section: Adaptive Wfqmentioning

confidence: 99%

See 3 more Smart Citations

Scheduling and Quality Differentiation in Differentiated Services

Antila¹,

Luoma²

2003

Interactive Multimedia on Next Generation Networks

View full text Add to dashboard Cite

show abstract

A Look-Ahead Scheduler to Provide Proportional Delay Differentiation in Wireless Network with a Multi-state Link

Lai

Chang

2003

Information Networking

View full text Add to dashboard Cite

A-Serv: a novel architecture providing scalable quality of service

Wang

Qian

Dauchy³

et al.

IEEE Global Telecommunications Conference, 2004. GLOBECOM '04.

View full text Add to dashboard Cite

QoS architectures define how routers process packets to ensure QoS service guarantees enforced. Existing QoS architectures such as Integrated Services (IntServ), Differentiated Services (DiffServ), and Dynamic Packet State (DPS) share one common property that the packet structure and the function of the routers are closely connected. Packets of one data flow are treated the same all the time at different routers. We propose to decouple such connection between packet structures and router functions. In our solution, packets carry as much information as possible, while routers process packets as detailed as possible until their load burden prohibits. We call such novel QoS architecture Adaptive Services (A-Serv). A-Serv utilizes our newly designed Load Adaptive Router to provide adaptive QoS to data flows. Treatments to data flows are not predefined but based on the load burden in Load Adaptive Routers. A-Serv overcomes the scalability problem of IntServ, provides better service guarantees to individual data flows than DiffServ and can be deployed incrementally. Our empirical analysis results show that compared with DiffServ architecture, A-Serv can provide differentiated services to data flows in the same DiffServ class and can provide better guaranteed QoS to data flows. Furthermore, AServ provides better protection to data flows than DiffServ when malicious data flows exist.

show abstract

A new scheduling mechanism to provide relative differentiation for real-time IP traffic

Cited by 11 publications

References 19 publications

Scheduling and Quality Differentiation in Differentiated Services

Scheduling and Quality Differentiation in Differentiated Services

A Look-Ahead Scheduler to Provide Proportional Delay Differentiation in Wireless Network with a Multi-state Link

A-Serv: a novel architecture providing scalable quality of service

Contact Info

Product

Resources

About