A Generalized Processor Sharing Approach to Flow Control in Integrated Service Networks:The Single Node Case.

A Generalized Processor Sharing Approach to Flow Control in Integrated Service Networks:The Single Node Case.Abhay.K.Parekh and Robert G.Gallager Laboratory for Information and Decision Systems Massachusetts Institute of TechnologyIEEE INFOCOM 1992

OutlineIntroductionMajor work of the paperGPS GPS ExamplePGPSDetermine the difference of delay and traffic in 2 scheme

OutlineBounds of buffer sizeVirtual System ImplementationLeaky-Bucket Admission ControlResultConclusionIntroductionThe Paper focus on a central problem in the control of congestion and flow in high speed integrated service networks.Goal is to find a implementable schemes for guaranteeing worst-case packet delay(Flow Control).paper packet scheduling Guarantee worst case packet delay4Major work of the paperMajor part of the paper is to provide a implementable scheme for guaranteeing worst-case packet delay.Show that PGPS ( Packet-by-Packet GPS ) combined with Leaky bucket admission control can achieve the goal. PGPS Leaky Bucket Admission Control worst-case packet delay weightsessios5Generalized Processor SharingGPS is a work-conserving flow control mechanism that ensure upper bounds of worst packet delay.But it can not be implemented because it assume that the packet size can be infinitely divided.Work Conserving server will not let bandwidth idle.GPS

GPS work conserving flow control mechanism

Work conserving session

GPS worst-case packet delay higher bound packet GPS

6Generalized Processor SharingGPS Server checks to see if a new source can be accommodated and, if so, takes actions (such as reserving transmission links or switching capacity) to ensure the quality of service desired.

GPS Server sessionservice qualityWorst-case session request session ( backlog )7Generalized Processor SharingOnce a source begins sending traffic, the network ensures that the agreed-upon values of traffic parameters are not violated.GPS ensure input-arrival patternAverage rate, buffer size, parameter8Generalized Processor SharingGPS Serer is defined asSi (,t) be the amount of session i traffic served in an time interval [,t]. Si (,t) / Sj (,t) i / j , j = 1,2,,N ,for any session i that is continuously backlogged in the time interval [,t].

GPS

Si (,t) ,t interval session I serve

GPS Si / Sj >=i / j session iweight > sesssion j weight , session Isession j 9Generalized Processor SharingBy Summering all session j ( in order to find session i s rate ), we can derive :Si (,t) j j (t-)r i ,where r is the processing rate of server and is the weight of the session.gi = i r / j j .

session j session irate = i r / j j .

10Generalized Processor SharingGPS is a attractive multiplexing scheme for a number of reasons:If the average rate of a session i less than gi the session can be guaranteed a throughput i ,independent of the demand of the other sessionThe delay of an arriving session i bit can be bounded as a function of the session I queue length , independent of the queues and arrivals of the other sessionsGPS:

sessionthroughput, session

Session delayqueue lengthfunction bound,sessionqueue length arrival pattern11Generalized Processor SharingBy varying i , we have flexibility to treat sessions in different way, as long as the total average rate of all sessions