OverQoS: An Overlay based Architecture for Enhancing Internet QoS Lakshminarayanan Subramanian, Ion Stoica, Hari Balakrishnan, and Randy H. Katz UCB and.

Slides:



Advertisements
Similar presentations
Streaming Video over the Internet
Advertisements

1 School of Computing Science Simon Fraser University CMPT 771/471: Internet Architecture & Protocols TCP-Friendly Transport Protocols.
CSIT560 Internet Infrastructure: Switches and Routers Active Queue Management Presented By: Gary Po, Henry Hui and Kenny Chong.
Engineering Internet QoS
Playback-buffer Equalization For Streaming Media Using Stateless Transport Prioritization By Wai-tian Tan, Weidong Cui and John G. Apostolopoulos Presented.
APOHN: Subnetwork Layering to Improve TCP Performance over Heterogeneous Paths April 4, 2006 Dzmitry Kliazovich, Fabrizio Granelli, University of Trento,
1 Service Differentiation at Transport Layer via TCP Westwood Low- Priority (TCPW-LP) H. Shimonishi, M.Y. Sanadidi and M. Geria System Platforms Research.
OverQoS: Offering QoS using Overlays Lakshmi Sahara Retreat, 2003.
1 Estimating Shared Congestion Among Internet Paths Weidong Cui, Sridhar Machiraju Randy H. Katz, Ion Stoica Electrical Engineering and Computer Science.
Shivkumar Kalyanaraman Rensselaer Polytechnic Institute 1 ECSE-6600: Internet Protocols Informal Quiz #07 Shivkumar Kalyanaraman: GOOGLE: “Shiv RPI”
The War Between Mice and Elephants Presented By Eric Wang Liang Guo and Ibrahim Matta Boston University ICNP
1 Sangeun Han, Athina Markopoulou Transmitting Scalable Video over a DiffServ network EE368C Project Proposal Sangeun Han, Athina Markopoulou 1/30/01.
Metrics for Performance Evaluation Nelson Fonseca State University of Campinas.
Internet Research Needs a Critical Perspective Towards Models –Sally Floyd –IMA Workshop, January 2004.
Yi Liang Multi-stream Voice Communication with Path Diversity.
Robust Scalable Video Streaming over Internet with Network-Adaptive Congestion Control and Unequal Loss Protection Quan Zang, Guijin Wang, Wenwu Zhu, and.
Predictive End-to-End Reservations via A Hierarchical Clearing House Endeavour Retreat June 19-21, 2000 Chen-Nee Chuah (Advisor: Professor Randy H. Katz)
Special Session PDCS’2000 Interworking of Diffserv, RSVP and MPLS for achieving QoS in the Internet Junaid Ahmed Zubairi Department of Mathematics and.
CS 268: Wireless Transport Protocols Kevin Lai Feb 13, 2002.
Katz, Stoica F04 EECS 122: Introduction to Computer Networks Packet Scheduling and QoS Computer Science Division Department of Electrical Engineering and.
Multiple Sender Distributed Video Streaming Thinh Nguyen, Avideh Zakhor appears on “IEEE Transactions On Multimedia, vol. 6, no. 2, April, 2004”
1 Traffic Sensitive Quality of Service Controller Masters Thesis Submitted by :Abhishek Kumar Advisors: Prof Mark Claypool Prof Robert Kinicki Reader:
Traffic Sensitive Active Queue Management - Mark Claypool, Robert Kinicki, Abhishek Kumar Dept. of Computer Science Worcester Polytechnic Institute Presenter.
1March -05 Jiangchuan Liu with Xinyan Zhang, Bo Li, and T.S.P.Yum Infocom 2005 CoolStreaming/DONet: A Data-Driven Overlay Network for Peer-to-Peer Live.
Yi Liang Aug. 2, 01 Low-latency Internet media streaming using packet path diversity - prior and future work.
1 Emulating AQM from End Hosts Presenters: Syed Zaidi Ivor Rodrigues.
1 Manpreet Singh, Prashant Pradhan* and Paul Francis * MPAT: Aggregate TCP Congestion Management as a Building Block for Internet QoS.
RRAPID: Real-time Recovery based on Active Probing, Introspection, and Decentralization Takashi Suzuki Matthew Caesar.
School of Information Technologies IP Quality of Service NETS3303/3603 Weeks
Building a Strong Foundation for a Future Internet Jennifer Rexford ’91 Computer Science Department (and Electrical Engineering and the Center for IT Policy)
Tiziana FerrariQuality of Service for Remote Control in the High Energy Physics Experiments CHEP, 07 Feb Quality of Service for Remote Control in.
PRISM: Proxies for Internet Streaming Media J. Kurose, P. Shenoy, D. Towsley (UMass/Amherst) L. Gao (Smith College) G. Hjalmtysson, J. Rexford (AT&T Research.
Providing Controlled Quality Assurance in Video Streaming across the Internet Yingfei Dong, Zhi-Li Zhang and Rohit Rakesh Computer Networking and Multimedia.
Advanced Network Architecture Research Group 2001/11/149 th International Conference on Network Protocols Scalable Socket Buffer Tuning for High-Performance.
End-to-end QoE Optimization Through Overlay Network Deployment Bart De Vleeschauwer, Filip De Turck, Bart Dhoedt and Piet Demeester Ghent University -
© 2006 Cisco Systems, Inc. All rights reserved. 3.3: Selecting an Appropriate QoS Policy Model.
Adaptive Packet Marking for Providing Differentiated Services in the Internet Wu-chang Feng, Debanjan Saha, Dilip Kandlur, Kang Shin October 13, 1998.
© 2006 Cisco Systems, Inc. All rights reserved. Optimizing Converged Cisco Networks (ONT) Module 3: Introduction to IP QoS.
Network management Reinhard Laroy BIPT European Parliament - 27 February 2012.
1 Dude, Where is My Packet?. NUS.SOC.CS5248 OOI WEI TSANG 2 Overview Characteristics of the Internet General techniques Error recovery for audio Effect.
RPT: Re-architecting Loss Protection for Content-Aware Networks Dongsu Han, Ashok Anand ǂ, Aditya Akella ǂ, and Srinivasan Seshan Carnegie Mellon University.
Beyond Best-Effort Service Advanced Multimedia University of Palestine University of Palestine Eng. Wisam Zaqoot Eng. Wisam Zaqoot November 2010 November.
Analytical Model of Hop-to-End based Network-Adaptive FEC scheme over Multi-hop Wireless Networks Koh Choi Networked Media Laboratory Dept. of.
Methods for providing Quality of Service in WLANs W.Burakowski, A. Beben, J.Sliwinski Institute of Telecommunications, Warsaw University of Technology,
TOMA: A Viable Solution for Large- Scale Multicast Service Support Li Lao, Jun-Hong Cui, and Mario Gerla UCLA and University of Connecticut Networking.
Making the Best of the Best-Effort Service (2) Advanced Multimedia University of Palestine University of Palestine Eng. Wisam Zaqoot Eng. Wisam Zaqoot.
Advanced Network Architecture Research Group 2001/11/74 th Asia-Pacific Symposium on Information and Telecommunication Technologies Design and Implementation.
Shivkumar Kalyanaraman Rensselaer Polytechnic Institute 1 Efficient Path Aggregation and Error Control for Video Streaming OMESH TICKOO, Shiv Kalyanaraman,
Multimedia Information System Lab. Network Architecture Res. Group MPEG-4 Video Transfer with TCP-Friendly Rate Control Naoki Wakamiya Osaka University,
Analysis of TCP Latency over Wireless Links Supporting FEC/ARQ-SR for Error Recovery Raja Abdelmoumen, Mohammad Malli, Chadi Barakat PLANETE group, INRIA.
Multimedia and QoS#1 Quality of Service Support. Multimedia and QoS#2 QOS in IP Networks r IETF groups are working on proposals to provide QOS control.
Improving TCP Performance over Wireless Networks
1 Wide Area Network Emulation on the Millennium Bhaskaran Raman Yan Chen Weidong Cui Randy Katz {bhaskar, yanchen, wdc, Millennium.
EE 122: Lecture 15 (Quality of Service) Ion Stoica October 25, 2001.
Low Latency Adaptive Streaming over TCP Authors Ashvin Goel Charles Krasic Jonathan Walpole Presented By Sudeep Rege Sachin Edlabadkar.
T. S. Eugene Ngeugeneng at cs.rice.edu Rice University1 COMP/ELEC 429 Introduction to Computer Networks Lecture 18: Quality of Service Slides used with.
OverQos: Less Bandwidth for More Reliable Service ?? --A Criticism Hongyu Gao Gregory Peaker.
SHADOWSTREAM: PERFORMANCE EVALUATION AS A CAPABILITY IN PRODUCTION INTERNET LIVE STREAM NETWORK ACM SIGCOMM CING-YU CHU.
1 On Maximum Rate Control of Weighted Fair Scheduling Jeng Farn Lee.
L Subramanian*, I Stoica*, H Balakrishnan +, R Katz* *UC Berkeley, MIT + USENIX NSDI’04, 2004 Presented by Alok Rakkhit, Ionut Trestian.
XCP: eXplicit Control Protocol Dina Katabi MIT Lab for Computer Science
1 Advanced Transport Protocol Design Nguyen Multimedia Communications Laboratory March 23, 2005.
Network Computing Laboratory Load Balancing and Stability Issues in Algorithms for Service Composition Bhaskaran Raman & Randy H.Katz U.C Berkeley INFOCOM.
Access Link Capacity Monitoring with TFRC Probe Ling-Jyh Chen, Tony Sun, Dan Xu, M. Y. Sanadidi, Mario Gerla Computer Science Department, University of.
MMPTCP: A Multipath Transport Protocol for Data Centres 1 Morteza Kheirkhah University of Edinburgh, UK Ian Wakeman and George Parisis University of Sussex,
PATH DIVERSITY WITH FORWARD ERROR CORRECTION SYSTEM FOR PACKET SWITCHED NETWORKS Thinh Nguyen and Avideh Zakhor IEEE INFOCOM 2003.
OverQos: An Overlay based Architecture for Enhancing Internet Qos L Subramanian*, I Stoica*, H Balakrishnan +, R Katz* *UC Berkeley, MIT + USENIX NSDI’04,
Muhammad Mateen Yaqoob Department of Computer Science COMSATS Institute of Information Technology, Abbottabad 1.
Project proposal Multi-stream and multi-path audio transmission
Presentation transcript:

OverQoS: An Overlay based Architecture for Enhancing Internet QoS Lakshminarayanan Subramanian, Ion Stoica, Hari Balakrishnan, and Randy H. Katz UCB and MIT NSDI 2004 Presented by Seungwoo Kang 2005/11/09

Korea Advanced Institute of Science and Technology Network Computing Laboratory 2/24 Contents Problem Approach OverQoS architecture Design principles Controlled-Loss Virtual Link Two sample applications Evaluation Critique

Korea Advanced Institute of Science and Technology Network Computing Laboratory 3/24 Problem Today’s Internet  Best-effort service –No guarantee for network delay or bw Efforts to provide QoS –Types Intserv Diffserv –Problems Difficulties of changing IP infrastructures –Underlying IP routers must be equipped with QoS mechanisms

Korea Advanced Institute of Science and Technology Network Computing Laboratory 4/24 Approach OverQoS –An overlay based QoS architecture to enhance Internet QoS –CLVL (Controlled-Loss Virtual Link) abstraction –QoS enhancements Smoothing losses Packet prioritization Statistical bandwidth and loss guarantee

Korea Advanced Institute of Science and Technology Network Computing Laboratory 5/24 OverQoS architecture Assumptions –Pre-determined placement of overlay nodes –Fixed end-to-end overlay path Terms –Virtual link –Bundle

Korea Advanced Institute of Science and Technology Network Computing Laboratory 6/24 Design principles Bundle loss control –Should bound the loss rate experienced by a bundle Resource management within a bundle –Control the loss and bw allocations of each flow and/or application

Korea Advanced Institute of Science and Technology Network Computing Laboratory 7/24 Design principles - Bundle Loss Control Bound on the loss rate –CLVL Recovering from network losses using FEC and ARQ Hybrid FEC/ARQ approach –Minimize the amount of redundancy traffic to meet the target loss rate, q c(t) = b(t) * (1 – r(t)) If arrival rate at the entry node < c(t), Packet loss rate, p will not exceed q, with high prob.

Korea Advanced Institute of Science and Technology Network Computing Laboratory 8/24 Design principles – Resource Management within a Bundle Statistical bw guarantee –Available bw, c, is time-varying –But, possible to provide a statistical bound of minimum bw, c min subset of the flows Empirical data –160Kbps, 269, 420 With 99% prob. P(c < c min ) = u

Korea Advanced Institute of Science and Technology Network Computing Laboratory 9/24 Controlled-Loss Virtual Link How to compute b, the maximum sending rate across an OverQoS link How to achieve q, the target loss rate for the flows in the bundle

Korea Advanced Institute of Science and Technology Network Computing Laboratory 10/24 Controlled-Loss Virtual Link - Estimating b Use MulTCP [ACM Comp. Comm. Review ‘98] –emulate the behavior of N TCP connections –Alpha = N/2, beta = 1/2N as the increment and decrement parameters

Korea Advanced Institute of Science and Technology Network Computing Laboratory 11/24 Controlled-Loss Virtual Link - Achieving target loss rate q Statistical bound on the avg. loss rate FEC and ARQ mechanism –Trade-off FEC: bw overhead ARQ: delay for recovery ARQ-based CLVL –Repeat the retransmission until success –How many times? L = log p q – 1, p: avg loss rate, q: target loss rate 10% loss 가 있는 상황에서 1% loss 를 맞추려면 1 번 더 보내야 지 ?

Korea Advanced Institute of Science and Technology Network Computing Laboratory 12/24 Controlled-Loss Virtual Link - Achieving target loss rate q FEC-based CLVL –Erasure code such as Reed-Solomon –(n, k), redundancy factor r = (n – k)/n FEC+ARQ based CLVL –# of retx : at most one Delay constraints for loss recovery –Better not to use FEC in the first round, use FEC only to protect retransmitted packets Minimize bw overhead

Korea Advanced Institute of Science and Technology Network Computing Laboratory 13/24 Node Architecture

Korea Advanced Institute of Science and Technology Network Computing Laboratory 14/24 Two sample applications Media streaming applications –Enhance media quality under lossy network condition ARQ-based CLVL Smoothing bursty losses Packet prioritization (e.g. I-frame, B, P) –Types Audio MPEG Multiplayer online game application –Prevent a skip or disconnection due to bursty losses –Counterstrike

Korea Advanced Institute of Science and Technology Network Computing Laboratory 15/24 Two sample applications – Media streaming applications M-K: 2%, I-L: 3% PESQ (Perceptual Evaluation of Speech Quality)

Korea Advanced Institute of Science and Technology Network Computing Laboratory 16/24 Two sample applications – Multiplayer online game application

Korea Advanced Institute of Science and Technology Network Computing Laboratory 17/24 Evaluation Items –Loss guarantee –Bw guarantee –OverQoS cost –Fairness/Stability Environment –Wide area testbed (RON, PlanetLab) 19 overlay nodes in diverse locations (Europe, Korea, Canada..) –Simulation (different types of traffic loss patterns) On top of ns-2 simulator Single link of 10 Mbps

Korea Advanced Institute of Science and Technology Network Computing Laboratory 18/24 Evaluation – Statistical loss guarantee Simulation Wide area evaluation –80 of the 83 VL the target q achieved based on FEC+ARQ based CLVL –Causes to fail – non-recoverable losses Short outages Bi-modal loss distribution

Korea Advanced Institute of Science and Technology Network Computing Laboratory 19/24 Evaluation – Statistical bw guarantee What bw guarantees are realizable on a virtual link? –c min > 100 Kbps for more than 80% of the links –At least 25% of the avg throughput in many cases –90 % of the avg throughput in some cases –Median value of c min /c avg : 0.4 and 0.35 for u = 0.01 and 0.005

Korea Advanced Institute of Science and Technology Network Computing Laboratory 20/24 Evaluation – Statistical bw guarantee

Korea Advanced Institute of Science and Technology Network Computing Laboratory 21/24 Evaluation – OverQoS cost Bw overheadDelay

Korea Advanced Institute of Science and Technology Network Computing Laboratory 22/24 Evaluation – Fairness and stability Ratio of throughputs of the three OverQoS bundles is preserved

Korea Advanced Institute of Science and Technology Network Computing Laboratory 23/24 Critiques Strong Points –Demonstration using real applications and extensive evaluation with real implementation and deployment Weak Points –IP router 를 고쳐야 하는 일은 없어졌으나 OverQoS 를 이용하기 위한 application proxy 를 만들어야 하는 일이 생겼다. 이런 일을 어떤 식으로 든 간편하게 만들어 주는 방법이 필요할 것이다. – 동일한 세팅에서 target loss 를 달성하는데 얼마만큼의 overhead 와 delay 가 발생하는지를 같이 보여주는 실험 결과가 있었으면.. – 가능한 Bw guarantee value 를 구하는데 그쳤는데 실제로 적용했을 때 어 떤 결과가 나왔을까를 보여줬으면.. –Media streaming application 테스트에서 delay variation 이 큰 문제가 없 다고 가정했는데 과연 적합한가 ? –Scalability issue: how many concurrent flows OverQoS can support?

End