DOC PREVIEW
WUSTL CIS 777 - ATM Traffic Management

This preview shows page 1-2-22-23 out of 23 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 23 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 23 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 23 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 23 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 23 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

Raj JainThe Ohio State University1Dollar Day SaleOne Megabit memory, One Megabyte disk,One Mbps link, One MIP processor, onedollar each.....ATM Traffic ManagementATM Traffic ManagementRaj JainProfessor of Computer and Info. SciencesThe Ohio State UniversityColumbus, OH 43210-1277http://www.cis.ohio-state.edu/~jain/Raj JainThe Ohio State University2Dollar Day SaleDollar Day SaleOne Megabit memory, One Megabyte disk, OneMbps link, One MIP processor, 10 cents each.....Raj JainThe Ohio State University3OverviewOverviewq Why worry about congestion?q Congestion schemes for ATMq Explicit Rate-based Controlq ABR Traffic ManagementRaj JainThe Ohio State University4FutureFutureIn 1990, the memory will be so cheap that you willnot have to worry about paging, swapping, virtualmemory, memory hierarchy, and....Year1980Raj JainThe Ohio State University5Why Worry About Congestion?Why Worry About Congestion?Q: Will the congestion problem be solved when:q Memory becomes cheap (infinite memory)?q Links become cheap (very high speed links)?q Processors become cheap?A: None of the above.No bufferOld age19.2 kb/s 1 Mb/sSSSSSSSSSSSSSSSSTime = 7 hoursFile transfer time = 5 minsRaj JainThe Ohio State University6Conclusions:q Congestion is a dynamic problem.Static solutions are not sufficientq Bandwidth explosion⇒ More unbalanced networksq Buffer shortage is a symptom not the cause.AABBSSCCDDAll links 1 Gb/sRaj JainThe Ohio State University7Economic ReasonsEconomic Reasonsq Network is a shared resourceBecause it is expensive and needed occasionally(Like airplanes, emergency rooms)q Most costs are fixed.Cost for fiber, switches, laying fiber and maintainingthem does not depend upon usage⇒ Underutilization is expensiveq But overutilization leads to user dissatisfaction.q Need a way to keep the network maximally utilizedRaj JainThe Ohio State University8Service CategoriesService CategoriesStandbyJoy RidersGuaranteedConfirmedRaj JainThe Ohio State University9Service CategoriesService Categoriesq ABR (Available bit rate):Source follows network feedback.Max throughput with minimum loss.q UBR (Unspecified bit rate):User sends whenever it wants. No feedback. Noguarantee. Cells may be dropped during congestion.q CBR (Constant bit rate): User declares required rate.Throughput, delay and delay variation guaranteed.q VBR (Variable bit rate): Declare avg and max rate.m rt-VBR (Real-time): Conferencing.Max delay guaranteed.m nrt-VBR (non-real time): Stored video.Raj JainThe Ohio State University10153246Traffic Management on theTraffic Management on theInformation SuperhighwayInformation SuperhighwayCACUPCSelectiveFrameDiscardShapingSchedulingTraffic Monitoring and feedback7Raj JainThe Ohio State University11Traffic Management FunctionsTraffic Management Functionsq Connection Admission Control (CAC): Can requestedbandwidth and quality of service be supported?q Traffic Shaping: Limit burst length. Space-out cells.q Usage Parameter Control (UPC):Monitor and control traffic at the network entrance.q Network Resource Management: Scheduling,Queueing, virtual path resource reservationq Selective cell discard:Cell Loss Priority (CLP) = 1 cells may be droppedCells of non-compliant connections may be droppedq Frame Discardingq Feedback ControlRaj JainThe Ohio State University12EFCIDestinationDestinationSourceSourceRMBinary Rate SchemeBinary Rate Schemeq DECbit scheme in many standards since 1986.q Forward explicit congestion notification (FECN) inFrame relayq Explicit forward congestion indicator (EFCI) set to 0at source. Congested switches set EFCI to 1q Every nth cell, destination sends an resourcemanagement (RM) cell to the sourceRaj JainThe Ohio State University13Go leftGo leftGo 30 km East35 km SouthGo 30 km East35 km SouthRaj JainThe Ohio State University14The Explicit Rate SchemeThe Explicit Rate Schemeq Sources send one RM cell every n cellsq The RM cells contain “Explicit rate”q Destination returns the RM cell to the sourceq The switches adjust the rate downq Source adjusts to the specified rateExplicit RateExplicit RateCurrent Cell RateCurrent Cell RateRaj JainThe Ohio State University15TTraffic Contract ParametersTraffic Contract Parametersq Peak Cell Rate (PCR): 1/Tq Cell Transfer Delay (CTD): First bit in to last bit outq Cell Delay Variation (CDV): ~ Max CTD - Min CTDm Peak-to-peak CDVq Cell Delay Variation Tolerance (CDVT) t= GCRA parameter wrt PCR Þ GCRA(T, t)Raj JainThe Ohio State University16q Sustained Cell Rate (SCR): Average over a longperiodq Burst Tolerance (BT) τs : GCRA parameter wrtSCR GCRA(1/Ts, τs)Maximum Burst Size:MBS = 1+BT/(1/SCR-1/PCR)BT ∈ [(MBS-1)(1/SCR-1/PCR), MBS(1/SCR-1/PCR)]q Cell Loss Ratio (CLR): Cells lost /Totals cells sentq Minimum cell rate (MCR)* Not negotiatedRaj JainThe Ohio State University17Peak-to-Peak CDVPeak-to-Peak CDVPeak-to-peak CDVFixedDelayMax CTDCell Transfer DelayProbabilityDensityRaj JainThe Ohio State University18Service CategoriesService Categories1Network specific2Not subject to CAC/UPC3PCR ⇒ Max ACR4Explicitly/implicitlyspecified for PVC/SVC5Not signaled. Different values mayapply at different interfaces alongthe path.6Follow ABR rulesAttribute CBR rt-VBR nrt-VBR UBR ABRPCR, CDVT4,5Specified Specified Specified Specified2Specified3SCR,MBS,CDVT4,5N/A Specified Specified N/A N/AMCR4N/A N/A N/A N/A SpecifiedPeak-to-peakCDVSpecified Specified Unspecified Unspecified UnspecifiedMax CTD Specified Specified Unspecified Unspecified UnspecifiedCLR4Specified Specified Specified Unspecified Specified1Feedback Unspecified Unspecified Unspecified Unspecified Specified6Raj JainThe Ohio State University19Segment-by-Segment ControlSegment-by-Segment Controlq Virtual source/virtual destinations follow allnotification/control rulesq Can be hop-by-hop❑ Virtual dest/sources maintain per-VC queues.SSDDEFCIRMEFCIRMEFCIRMEFCIRMSSDDEFCIRMEFCIRMRaj JainThe Ohio State University20Congestion: SummaryCongestion: Summaryq Traffic Management is key tosuccess of ATMq Several different methods: CAC,Shaping, UPC, Scheduling, ...q Service categories:CBR, VBR,ABR, UBRq ER switches provide much betterperformance than EFCI.Raj JainThe Ohio State University21HomeworkHomeworkq Read Chapter 16 of Stallings’ bookRaj JainThe Ohio State University22ReferencesReferencesq R.Jain, "Congestion Control and TrafficManagement in ATM Networks: Recent Advancesand A Survey", Computer Networks and ISDNSystems, November 1996, http://www.cis.ohio-state.edu/~jain/q K. Siu and R. Jain, "A


View Full Document

WUSTL CIS 777 - ATM Traffic Management

Documents in this Course
SONET

SONET

17 pages

SONET

SONET

22 pages

SONET

SONET

19 pages

Load more
Download ATM Traffic Management
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 ATM Traffic Management 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 ATM Traffic Management 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?