1 Value of Supporting Class-of- Service in IP Backbones Murat Yuksel (University of Nevada – Reno) K. K. Ramakrishnan (AT&T Labs Research)

Slides:



Advertisements
Similar presentations
June 27, 2007 FIND Meeting, From Packet-Switching to Contract- Switching Aparna Gupta Shivkumar Kalyanaraman Rensselaer Polytechnic Institute Troy,
Advertisements

FIND PI Meeting, April Contract-Switching: Value Flows in Inter-Domain Routing Murat Yuksel University of Nevada – Reno Reno, NV Aparna Gupta, Koushik.
1 Praveen K. Muthuswamy Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute In collaboration with Koushik Kar, Aparna Gupta (RPI)
Traffic and routing. Network Queueing Model Packets are buffered in egress queues waiting for serialization on line Link capacity is C bps Average packet.
Quantifying Overprovisioning vs. Class-of-Service: Informing the Net Neutrality Debate Murat Yuksel (University of Nevada – Reno) K.
Playback-buffer Equalization For Streaming Media Using Stateless Transport Prioritization By Wai-tian Tan, Weidong Cui and John G. Apostolopoulos Presented.
CNDS 2001, Phoenix, AZ Simulating the Smart Market Pricing Scheme on Differentiated- Services Architecture Murat Yuksel and Shivkumar Kalyanaraman Rensselaer.
Bilal Gonen University of Alaska Anchorage Murat Yuksel University of Nevada, Reno.
Competitive Analysis of Buffer Policies with SLA Commitments Boaz Patt-Shamir, Tel Aviv University Gabriel Scalosub, University of Toronto Yuval Shavitt,
SIGCOMM 2003 Making Intra-Domain Routing Robust to Changing and Uncertain Traffic Demands: Understanding Fundamental Tradeoffs David Applegate Edith Cohen.
Comments on the Performance of Measurement Based Admission Control Algorithms Lee Breslau, S. Jamin, S. Shenker Infocom 2000.
© 2007 AT&T Knowledge Ventures. All rights reserved. AT&T and the AT&T logo are trademarks of AT&T Knowledge Ventures. The Evolving Internet Traffic, Engineering,
Fair queueing and congestion control Jim Roberts (France Telecom) Joint work with Jordan Augé Workshop on Congestion Control Hamilton Institute, Sept 2005.
Network Architecture for Joint Failure Recovery and Traffic Engineering Martin Suchara in collaboration with: D. Xu, R. Doverspike, D. Johnson and J. Rexford.
Nick McKeown CS244 Lecture 6 Packet Switches. What you said The very premise of the paper was a bit of an eye- opener for me, for previously I had never.
Designing Networks with Little or No Buffers or Can Gulliver Survive in Lilliput? Yashar Ganjali High Performance Networking Group Stanford University.
High Performance All-Optical Networks with Small Buffers Yashar Ganjali High Performance Networking Group Stanford University
Congestion Pricing Overlaid on Edge-to-Edge Congestion Control Murat Yuksel, Shivkumar Kalyanaraman and Anuj Goel Rensselaer Polytechnic Institute, Troy,
Pricing Granularity for Congestion-Sensitive Pricing Murat Yüksel and Shivkumar Kalyanaraman Rensselaer Polytechnic Institute, Troy, NY {yuksem, shivkuma}
Traffic Engineering Jennifer Rexford Advanced Computer Networks Tuesdays/Thursdays 1:30pm-2:50pm.
Traffic Forecasting Medium Access TRANSFORMA Vladislav Petkov Katia Obraczka 1.
Shivkumar Kalyanaraman Rensselaer Polytechnic Institute 1 Introduction to Experiment Design Shiv Kalyanaraman Rensselaer Polytechnic Institute
Modeling TCP in Small-Buffer Networks
Current Research Topics -Sigcomm Sessions -QoS -Network analysis & security -Multicast -giga/tera bit routers /fast classification -web performance -TCP.
Reducing the Buffer Size in Backbone Routers Yashar Ganjali High Performance Networking Group Stanford University February 23, 2005
A novel switching paradigm for buffer-less WDM networks Myungsik Yoo and Chunming Qiao EE and CSE Departments University at Buffalo (SUNY)
A Strategy for Implementing Smart Market Pricing Scheme on Diff-Serv Murat Yuksel and Shivkumar Kalyanaraman Rensselaer Polytechnic Institute, Troy, NY.
Distributed-Dynamic Capacity Contracting: A congestion pricing framework for Diff-Serv Murat Yuksel and Shivkumar Kalyanaraman Rensselaer Polytechnic Institute,
Trajectory-Based Forwarding Mechanisms for Ad-Hoc Sensor Networks Murat Yuksel, Ritesh Pradhan, Shivkumar Kalyanaraman Electrical, Computer, and Systems.
Statistical Approach to NoC Design Itamar Cohen, Ori Rottenstreich and Isaac Keslassy Technion (Israel)
Elasticity Considerations for Optimal Pricing of Networks Murat Yüksel and Shivkumar Kalyanaraman Rensselaer Polytechnic Institute, Troy, NY {yuksem, shivkuma}
Rensselaer Polytechnic Institute © Shivkumar Kalvanaraman & © Biplab Sikdar1 ECSE-4730: Computer Communication Networks (CCN) Network Layer Performance.
September 12, 2006IEEE PIMRC 2006, Helsinki, Finland1 On the Packet Header Size and Network State Tradeoff for Trajectory-Based Routing in Wireless Networks.
EECC694 - Shaaban #1 lec #7 Spring The OSI Reference Model Network Layer.
10th Workshop on Information Technologies and Systems 1 A Comparative Evaluation of Internet Pricing Schemes: Smart Market and Dynamic Capacity Contracting.
Internet Queuing Delay Introduction How many packets in the queue? How long a packet takes to go through?
MATE: MPLS Adaptive Traffic Engineering Anwar Elwalid, et. al. IEEE INFOCOM 2001.
Roadmap-Based End-to-End Traffic Engineering for Multi-hop Wireless Networks Mustafa O. Kilavuz Ahmet Soran Murat Yuksel University of Nevada Reno.
IEEE Global Internet, April Contract-Switching Paradigm for Internet Value Flows and Risk Management Murat Yuksel University.
IEEE LANMAN, Cluj, Romania, September Cross-Layer Failure Restoration for a Robust IPTV Service Murat Yuksel
Tomo-gravity Yin ZhangMatthew Roughan Nick DuffieldAlbert Greenberg “A Northern NJ Research Lab” ACM.
1 Chapters 9 Self-SimilarTraffic. Chapter 9 – Self-Similar Traffic 2 Introduction- Motivation Validity of the queuing models we have studied depends on.
Cost-Performance Tradeoffs in MPLS and IP Routing Selma Yilmaz Ibrahim Matta Boston University.
Using Dijkstra’s Algorithm to Find a Shortest Path from a to z 1.
L13. Shortest path routing D. Moltchanov, TUT, Spring 2008 D. Moltchanov, TUT, Spring 2014.
1 Kommunikatsiooniteenuste arendus IRT0080 Loeng 7 Avo Ots telekommunikatsiooni õppetool, TTÜ raadio- ja sidetehnika inst.
“Intra-Network Routing Scheme using Mobile Agents” by Ajay L. Thakur.
Budapest University of Technology and Economics Department of Telecommunications and Media Informatics Optimized QoS Protection of Ethernet Trees Tibor.
Computer Networks Performance Metrics. Performance Metrics Outline Generic Performance Metrics Network performance Measures Components of Hop and End-to-End.
Understanding the Performance of TCP Pacing Amit Aggarwal, Stefan Savage, Thomas Anderson Department of Computer Science and Engineering University of.
A Case Study in Understanding OSPFv2 and BGP4 Interactions Using Efficient Experiment Design David Bauer†, Murat Yuksel‡, Christopher Carothers† and Shivkumar.
Shivkumar Kalyanaraman Rensselaer Polytechnic Institute 1 A TCP Friendly Traffic Marker for IP Differentiated Services Feroz Azeem, Shiv Kalyanaraman,
1 Module 4: Implementing OSPF. 2 Lessons OSPF OSPF Areas and Hierarchical Routing OSPF Operation OSPF Routing Tables Designing an OSPF Network.
Shivkumar Kalyanaraman Rensselaer Polytechnic Institute 1 Efficient Path Aggregation and Error Control for Video Streaming OMESH TICKOO, Shiv Kalyanaraman,
Queuing Theory and Traffic Analysis Based on Slides by Richard Martin.
Multiplicative Wavelet Traffic Model and pathChirp: Efficient Available Bandwidth Estimation Vinay Ribeiro.
Florida State UniversityZhenhai Duan1 BCSQ: Bin-based Core Stateless Queueing for Scalable Support of Guaranteed Services Zhenhai Duan Karthik Parsha Department.
N. Hu (CMU)L. Li (Bell labs) Z. M. Mao. (U. Michigan) P. Steenkiste (CMU) J. Wang (AT&T) Infocom 2005 Presented By Mohammad Malli PhD student seminar Planete.
On Selfish Routing In Internet-like Environments Lili Qiu (Microsoft Research) Yang Richard Yang (Yale University) Yin Zhang (AT&T Labs – Research) Scott.
1 Presented by Sarbagya Buddhacharya. 2 Increasing bandwidth demand in telecommunication networks is satisfied by WDM networks. Dimensioning of WDM networks.
1 Lecture 17 Fairness and Neutrality University of Nevada – Reno Computer Science & Engineering Department Fall 2015 CS 791 Special Topics: Network Architectures.
1 Routing and Resilience in Future Optical Broadband Telecommunications Networks 21 st January 2004 Andrew S. T. Lee Supervisor: Dr. David Harle Broadband.
Distance Vector Routing
1 Traffic Engineering By Kavitha Ganapa. 2 Introduction Traffic engineering is concerned with the issue of performance evaluation and optimization of.
Tango1 Considering End-to-End QoS Constraints in IP Network Design and Planning M.Ajmone Marsan, M. Garetto, E. Leonardi. M. Mellia, E. Wille Dipartimento.
Designing Multi-hop Wireless Backhaul Networks with Delay Guarantees Girija Narlikar, Gordon Wilfong, and Lisa Zhang Bell Lab. Infocom 2006.
Interaction and Animation on Geolocalization Based Network Topology by Engin Arslan.
Multi-Core Parallel Routing
Elasticity Considerations for Optimal Pricing of Networks
QoS routing Finding a path that can satisfy the QoS requirement of a connection. Achieving high resource utilization.
Presentation transcript:

1 Value of Supporting Class-of- Service in IP Backbones Murat Yuksel (University of Nevada – Reno) K. K. Ramakrishnan (AT&T Labs Research) Shiv Kalyanaraman (Rensselaer Polytechnic Institute) Joseph D. Houle (AT&T) Rita Sadhvani (AT&T)

2 Motivation: Thick (Over-provisioned) or Thin (Engineered) Pipes ? Thin: How to deal with bursts/overload? And meet premium SLAs… ! Thick: Cost of overprovisioning? Can this commodity model break even? rate time [Jim Roberts et al.] Media-rich applications require performance guarantees: e.g.: VoIP requires <300ms round-trip delay, <1% loss How to respond to these application needs? CoS approach: provide priority (i.e. higher class) to premium traffic Classless (best-effort) service approach: over-provision the capacity Question: How much extra capacity does the classless service require to match the performance of the higher class (premium) service in the CoS approach? rate time

3 Two Service Types: CoS vs. Classless Premium BE DD CoS Link (differentiated) D Prem = g D BE =(1-g) D D GIVEN:  D, D and a performance target (i.e. t target or p target ) FIND: What is the minimum  N that gives the same performance as in the premium class of the CoS case?  N =? Classless Link (neutral) BE Scheduling (e.g. priority)

4 REC: Required Extra Capacity REC= - =  N -  D (rate) = 100(  N /  D – 1)(%) How to quantify REC?

5 Link Model: Poisson traffic Assume: Poisson traffic, Exponential packet lengths for traffic in each class i.e. Premium class traffic is Poisson with g D Best-effort class traffic is Poisson with (1-g) D The aggregate traffic for the neutral link is also Poisson with rate D conservative: the superposition would be more bursty Delay: M/M/1  N = 1/t target + D Loss: M/M/1/K

6 More Bursty Traffic: MMPP MMPP = Markov-Modulated Poisson Process Easy to do the math… Simplest MMPP is of two states. MMPP traffic with mean D Traffic w/ equivalent rate to the neutral case, but w/ more burstiness. 1 2 Higher r means more bursty traffic.

7 Simulated Link Model: Delay MMPP/M/1 model a=0.5, r=8 a=0.5, r=4 If packet size is 1KB and the CoS link is  D = 10Gb/s: 5,000packets of delay = 4.1 ms REC can be quite high even for very small g and medium utilization.

8 a=0.5, r=4, K=15 Simulated Link Model: Loss MMPP/M/1/K model a=0.5, r=4, K=6 For a 1Gb/s link carrying 1KB packets: K = ~6pkts  0.1ms buffer time K = ~15pkts  0.25ms buffer time K = ~60pkts  1ms buffer time a=0.5, r=4, K=60

9 Network Model Steps to calculate REC for a network: Step 1: Construct the routing matrix R FxL based on shortest path Run Dijkstra’s algorithm on the topology matrices A NxN and W NxN Step 2: Form the traffic vector Fx1 from T NxN Step 3: Calculate the traffic load on each link: R T = Q Step 4: Check the feasibility of the traffic load and routing For any link If link capacity is less than the traffic load (e.g. C < Q) then update T accordingly and go to Step 2. Step 5: Calculate the required per-link REC (i.e.  N -  D ) by using Q I as the traffic rate D for Ith link, and the performance goal p target or t target. Used Rocketfuel topologies for A NxN and W NxN. Used gravity model for T NxN. Made a look- up to the simulated link model results.

10 Network Model: Delay Conservative MMPP: a=0.5, r=4 Abovenet Sprintlink Queuing delay range for legacy applications such as VoIP.

11 Network Model: Loss Conservative MMPP: a=0.5, r=4 Sprintlink Abovenet Reasonable buffer: K=60 pkts Loss probability range for typical ISP practices.

12 Summary A framework to study REC for delay or loss being the performance target. Link model REC grows when: traffic becomes more bursty the utilization of the CoS link becomes higher the performance target becomes tighter the fraction g of the Premium class traffic becomes smaller With conservative burstiness assumptions of MMPP traffic, REC ranges up to 100% even when g is 0.2 and the CoS link utilization is 40%. Network model: For legacy g2g performance targets, REC ranges from 50% to over 100% as g reduces below 0.5 and the CoS link utilization goes up to 60%.

13 Thank you! THE END