DOC PREVIEW
Purdue CS 59000 - A Measurement Based Admission Control Algorithm

This preview shows page 1-2-3-4-5-6 out of 18 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 18 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 18 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 18 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 18 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 18 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 18 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 18 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

Slide 1Types of ServicesSlide 3Slide 4Objective of paperMeasurement Based Admission ControlMeasurement Based Admission ControlWorst case delayComputation of new delay boundEquivalent Token bucket filterAdmission control algorithmSlide 12Measurement mechanism: Time windowMeasuring delayMeasuring ratePerformance TuningCOPSSlide 18A Measurement Based Admission Control Algorithm for Integrated Services Packet Networks ( extended version)S. Jamin, P. Danzig, S. Shenker, L. Zhang2Types of ServicesGuaranteed service a absolute bound on delay of every packet is providedPredictive service a fair but not absolute bound on delay of every packet is provided3Guaranteed serviceProvides hard or absolute bound on delay of every packetSources are described by either peak or average rates or filter like token bucketDescriptions provide upper bounds on traffic generated by sourceAdmission control algorithms use these a priori characterizations to calculate worst case behavior of flowsNetwork utilization low when flows are burstyNetwork utilization acceptable when flows are smooth4Predictive ServiceDoes not provide absolute bound on delayservice commitment is somewhat less reliablesources characterized by token bucket filters at admission timebehavior of existing flows determined by measurement rather than a priori characterizations when only few flows present, unpredictability of individual flows causes measurement based approach to be conservativedelivers significant gain in utilization only when there is high degree of statistical multiplexing5Objective of paperDescribes a measurement based admission control algorithm for predictive serviceTo answer the following questions can reliable delay bounds be provided for measurement based admission control algorithm does predictive service increase network utilization6Measurement Based Admission ControlSet of criteria controlling whether to admit a new flow based on approximate model of traffic flows uses measured quantities as inputs Measurement process to measure the inputsset of criteria used in this case  delay bound as a result of admitting the new flow link utilization as a result of admitting the new flow7Measurement Based Admission ControlSources requesting services characterize worst case behaviorto guarantee hard bound on delay worst case behavior considered conservative estimate of bound obtained new flow may not be admitted due to conservative estimateresults in low utilizationto achieve fairly reliable bound for predictive service approximate maximal delay by replacing worst case parameters with measured quantities results in higher utilization8Worst case delayTheorem by Parekh says the delay bound for a class j is the one time delay burst that accrues if the aggregate bucket of all classes 1 through j flows are simultaneously dumped into the switch and all classes 1 through j-1 sources continue to send at their reserved rates with j having lowest prioritythis worst case delay can change when flow of same class is admitted when flow of higher priority class is admitted when guaranteed flow is admittedclasses classified according to the delay bounds tolerated by that class9Computation of new delay boundThe delay bounds as a result of admitting the new flow is computedfor e.g., When new predictive flow  of same class is admitted new delay bound = old delay bound + k( bucket size of similarly new delay bounds are calculated for the caseWhen new predictive flow  of lower class is admittedWhen new guaranteed flow  is admitted10Equivalent Token bucket filterIn equations replace worst case values by measured valuessubstitute the reserved rates by the measured ratessubstitute the worst case delays by the measured delaysthis results in describing the existing aggregate traffic by an equivalent token bucket filter with parameters determined from traffic measurement11Admission control algorithmNew predictive flow: if incoming flow requests predictive service in a particular class request denied if sum of flow’s requested rate and current usage exceeds link utilization levelrequest denied if admitting new flow violates delay bound of the same priority levelrequest denied if admitting new flow violates delay bound of the lower priority level12Admission control algorithmNew guaranteed flow: if incoming flow requests guaranteed service request denied if total bandwidth of the flows including new flow exceeds link utilization levelrequest denied if admitting new guaranteed flow violates delay bound of predictive services this is due to the decrease in the bandwidth available for predictive flows due to admission of a guaranteed flow13Measurement mechanism: Time windowTwo measurementsexperienced delayutilizationto estimate delays measure the queuing delays d of every packetto estimate utilization sample the usage rate of the guaranteed service and the predictive class over sampling period of S packet transmission units14Measuring delayMeasurement variable to estimate queuing delay is DMeasurement window of T packet transmission units used as basic measurement blockvalue of D updated on three occasions  at end of measurement block D updated to reflect maximal packet delay seen in previous blockwhen individual delay measurement exceeds estimated maximum queuing delay D is updated to times the sampled delaywhenever a new flow is admitted D is updated to the value of projected delay from equationλ15Measuring rateMeasurement variable to estimate queuing delay is VMeasurement window of T packet transmission units used as basic measurement blockvalue of V updated on three occasions  at end of measurement block V updated to reflect maximal sampled utilization seen in previous blockwhen individual utilization measurement exceeds estimated V, V is updated to the sampled valuewhenever a new flow is admitted V is updated16Performance Tuning Parameters , S, T can be varied for better performancesmaller S, more sensitive to bursts , larger S smoother trafficsmaller T means more adaptability to changes in traffic load, larger T results in more stability larger T results in fewer delay violations and lower link utilization λ17COPS A simple client/server model for


View Full Document

Purdue CS 59000 - A Measurement Based Admission Control Algorithm

Documents in this Course
Lecture 4

Lecture 4

42 pages

Lecture 6

Lecture 6

38 pages

Load more
Download A Measurement Based Admission Control Algorithm
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view A Measurement Based Admission Control Algorithm and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view A Measurement Based Admission Control Algorithm 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?