Confused, Timid and Unstable: Picking a Video Rate is Hard Te-Yuan (TY) Huang Stanford University Nov 15 th, 2012 Joint work with Nikhil Handigol, Brandon.

Slides:



Advertisements
Similar presentations
Junchen Jiang (CMU) Vyas Sekar (Stony Brook U)
Advertisements

Introduction 2 1: Introduction.
IEEE JSAC Special Issue Adaptive Media Streaming Submissions by April 1 Details at
Saamer Akhshabi, Constantine Dovrolis Georgia Institute of Technology An Experimental Evaluation of Rate Adaptation Algorithms in Adaptive Video Streaming.
Confused, Timid, and Unstable: Picking a Video Streaming Rate is Hard Published in 2012 ACM’s Internet Measurement Conference (IMC) Five students from.
Doc.: IEEE /0604r1 Submission May 2014 Slide 1 Modeling and Evaluating Variable Bit rate Video Steaming for ax Date: Authors:
1 CS492B Project #2 TCP Tutorial # Jin Hyun Ju.
Winter 2008CS244a Handout #61 CS244a: An Introduction to Computer Networks Handout 6: The Transport Layer, Transmission Control Protocol (TCP), and User.
Ahmed Mansy, Mostafa Ammar (Georgia Tech) Bill Ver Steeg (Cisco)
Improving TCP Performance over MANETs by Exploiting Cross-Layer Information Awareness Xin Yu NYU Presented by: David Choffnes.
Performance Interactions Between P-HTTP and TCP Implementations J. Heidemann ACM Computer Communication Review April 1997 김호중 CA Lab., KAIST.
Measurements of Congestion Responsiveness of Windows Streaming Media (WSM) Presented By:- Ashish Gupta.
The Tension Between High Video Rate and No Rebuffering Te-Yuan (TY) Huang Stanford University IRTF Open 87 July 30th, 2013 Joint work Prof.
1 Modeling and Emulation of Internet Paths Pramod Sanaga, Jonathon Duerig, Robert Ricci, Jay Lepreau University of Utah.
1 Lecture 10: TCP Performance Slides adapted from: Congestion slides for Computer Networks: A Systems Approach (Peterson and Davis) Chapter 3 slides for.
High Performance Networking with Little or No Buffers Yashar Ganjali High Performance Networking Group Stanford University
High Performance Networking with Little or No Buffers Yashar Ganjali on behalf of Prof. Nick McKeown High Performance Networking Group Stanford University.
Sizing Router Buffers Nick McKeown Guido Appenzeller & Isaac Keslassy SNRC Review May 27 th, 2004.
Congestion Control in Distributed Media Streaming Lin Ma Wei Tsang Ooi School of Computing National University of Singapore IEEE INFOCOM 2007.
Reducing the Buffer Size in Backbone Routers Yashar Ganjali High Performance Networking Group Stanford University February 23, 2005
1 Emulating AQM from End Hosts Presenters: Syed Zaidi Ivor Rodrigues.
Isaac Keslassy (Technion) Guido Appenzeller & Nick McKeown (Stanford)
Medium Start in TCP-Friendly Rate Control Protocol CS 217 Class Project Spring 04 Peter Leong & Michael Welch.
The War Between Mice and Elephants By Liang Guo (Graduate Student) Ibrahim Matta (Professor) Boston University ICNP’2001 Presented By Preeti Phadnis.
Stanford University August 22, 2001 TCP Switching: Exposing Circuits to IP Pablo Molinero-Fernández Nick McKeown Stanford University.
1 K. Salah Module 6.1: TCP Flow and Congestion Control Connection establishment & Termination Flow Control Congestion Control QoS.
Reducing Flow-Completion Time for Small Flows by Modifying Slow-Start Affan Rauf ( )
03/12/08Nuova Systems Inc. Page 1 TCP Issues in the Data Center Tom Lyon The Future of TCP: Train-wreck or Evolution? Stanford University
Junxian Huang 1 Feng Qian 2 Yihua Guo 1 Yuanyuan Zhou 1 Qiang Xu 1 Z. Morley Mao 1 Subhabrata Sen 2 Oliver Spatscheck 2 1 University of Michigan 2 AT&T.
- Conviva Confidential - Understanding and Improving Video Quality Vyas Sekar, Ion Stoica, Hui Zhang.
All rights reserved © 2006, Alcatel Accelerating TCP Traffic on Broadband Access Networks  Ing-Jyh Tsang 
Courtesy: Nick McKeown, Stanford 1 TCP Congestion Control Tahir Azim.
Advanced Network Architecture Research Group 2001/11/149 th International Conference on Network Protocols Scalable Socket Buffer Tuning for High-Performance.
Transport Layer3-1 Chapter 3 outline r 3.1 Transport-layer services r 3.2 Multiplexing and demultiplexing r 3.3 Connectionless transport: UDP r 3.4 Principles.
Transport Layer3-1 Chapter 3 outline r 3.1 Transport-layer services r 3.2 Multiplexing and demultiplexing r 3.3 Connectionless transport: UDP r 3.4 Principles.
CS144 An Introduction to Computer Networks
1 Modeling the Effect of a Rate Smoother on TCP Congestion Control Behavior Kang Li, Jonathan Walpole, David C. Steere {kangli, walpole,
COMT 4291 Communications Protocols and TCP/IP COMT 429.
CS540/TE630 Computer Network Architecture Spring 2009 Tu/Th 10:30am-Noon Sue Moon.
An Efficient Approach for Content Delivery in Overlay Networks Mohammad Malli Chadi Barakat, Walid Dabbous Planete Project To appear in proceedings of.
Understanding the Performance of TCP Pacing Amit Aggarwal, Stefan Savage, Thomas Anderson Department of Computer Science and Engineering University of.
B 李奕德.  Abstract  Intro  ECN in DCTCP  TDCTCP  Performance evaluation  conclusion.
Parameswaran, Subramanian
Making the Best of the Best-Effort Service (2) Advanced Multimedia University of Palestine University of Palestine Eng. Wisam Zaqoot Eng. Wisam Zaqoot.
27th, Nov 2001 GLOBECOM /16 Analysis of Dynamic Behaviors of Many TCP Connections Sharing Tail-Drop / RED Routers Go Hasegawa Osaka University, Japan.
Transport Layer3-1 Announcements r Collect homework r New homework: m Ch3#3,4,7,10,12,16,18-20,25,26,31,33,37 m Due Wed Sep 24 r Reminder: m Project #1.
Advanced Network Architecture Research Group 2001/11/74 th Asia-Pacific Symposium on Information and Telecommunication Technologies Design and Implementation.
HighSpeed TCP for High Bandwidth-Delay Product Networks Raj Kettimuthu.
An Experimental Evaluation of Voice Quality over the Datagram Congestion Control Protocol H. Balan International Univeristy Bremen L. Eggert Nokia Research.
TCP Trunking: Design, Implementation and Performance H.T. Kung and S. Y. Wang.
Transport Layer3-1 TCP throughput r What’s the average throughout of TCP as a function of window size and RTT? m Ignore slow start r Let W be the window.
Self-generated Self-similar Traffic Péter Hága Péter Pollner Gábor Simon István Csabai Gábor Vattay.
1 SIGCOMM ’ 03 Low-Rate TCP-Targeted Denial of Service Attacks A. Kuzmanovic and E. W. Knightly Rice University Reviewed by Haoyu Song 9/25/2003.
Performance Interactions Between P-HTTP and TCP Implementation John Heidemann USC/Information Sciences Institute May 19, 1997 Presentation Baekcheol Jang.
Chapter 11.4 END-TO-END ISSUES. Optical Internet Optical technology Protocol translates availability of gigabit bandwidth in user-perceived QoS.
Development of a QoE Model Himadeepa Karlapudi 03/07/03.
79 Sidevõrgud IRT 4060/ IRT 0020 vooruloeng 8 / 3. nov 2004 Vooülekanne Avo Ots telekommunikatsiooni õppetool, TTÜ raadio- ja sidetehnika inst.
Matlab Tutorial: Netflix Trace Analysis (material available at 1 Dr. Jim Martin Associate Professor.
An Analysis of AIMD Algorithm with Decreasing Increases Yunhong Gu, Xinwei Hong, and Robert L. Grossman National Center for Data Mining.
Performance Evaluation of Redirection Schemes in Content Distribution Networks Jussi Kangasharju, Keith W. Ross Institut Eurecom Jim W. Roberts France.
Buffers: How we fell in love with them, and why we need a divorce Hot Interconnects, Stanford 2004 Nick McKeown High Performance Networking Group Stanford.
Accelerating Peer-to-Peer Networks for Video Streaming
TCP-LP: A Distributed Algorithm for Low Priority Data Transfer
Mohammad Malli Chadi Barakat, Walid Dabbous Alcatel meeting
SCTP v/s TCP – A Comparison of Transport Protocols for Web Traffic
TCP-LP: A Distributed Algorithm for Low Priority Data Transfer
Available Bit Rate Streaming
ECF: an MPTCP Scheduler to Manage Heterogeneous Paths
CSE679: Multimedia and Networking
Modeling and Evaluating Variable Bit rate Video Steaming for ax
Presentation transcript:

Confused, Timid and Unstable: Picking a Video Rate is Hard Te-Yuan (TY) Huang Stanford University Nov 15 th, 2012 Joint work with Nikhil Handigol, Brandon Heller, Nick McKeown, Ramesh Johari

Video is the BIG thing on the Internet Video is more than 50% of peak traffic in the US Trend: Streaming over HTTP – Content Distribution Networks (CDNs) Well-provisioned HTTP servers at the edge of the Internet Cheap (2-3 cents per GB in 2011) – Firewall friendliness 2Te-Yuan (TY) ACM IMC'12, Boston

3

4 Before download started After download started Te-Yuan (TY) ACM IMC'12, Boston

What happened? Both the download and video are over HTTP – TCP shares my home link equally among all flows 5 What is the problem? Te-Yuan (TY) ACM IMC'12, Boston

Experiment Setup The Internet Bandwidth Controller CDN 1 CDN 2 CDN 3 6 Content Distribution Networks Te-Yuan (TY) ACM IMC'12, Boston

Video Rate in the Presence of a Competing Flow Video Rate 7 Available Video Rates Fair share Competing Flow Throughout Steps down all the way to the lowest quality Video Flow Throughput Te-Yuan (TY) ACM IMC'12, Boston

What If We Manually Select a Video Rate? 8 Available Video Rates Manually select 1750kb/s Video Flow Throughput Video keeps playing without any problem Video Rate Te-Yuan (TY) ACM IMC'12, Boston

What If We Manually Select a Video Rate? 9 Competing Flow Throughput Video Rate Available Video Rates Manually select 1750kb/s Competing flow throughput is brought down Video Flow Throughput Te-Yuan (TY) ACM IMC'12, Boston

Not Just One Service’s Problem This happens in all the services we measured – Hulu, Netflix, Vudu 10 Video Rate Te-Yuan (TY) ACM IMC'12, Boston

The Problem 11 Why? It picks a video rate that is much too low Video client ends up with much less throughput than its fair share Te-Yuan (TY) ACM IMC'12, Boston

Outline How does rate selection over HTTP work? The rate selection process Where does it go wrong? The complete story Conclusion 12Te-Yuan (TY) ACM IMC'12, Boston

How Does Rate Selection over HTTP Work? The Internet Client 13 Standard, commoditized HTTP Servers Rate selection logic resides at client side CDN 3 File 1: File 2: Get File 1 (1750kb/s) Serve the video with quality 1750kb/s Playout Buffer Te-Yuan (TY) ACM IMC'12, Boston

The Rate Selection Process Initial video rate Video rate for the next video segment 14 Bandwidth estimation Download & measure Pick a rate What goes wrong? Te-Yuan (TY) ACM IMC'12, Boston

The Rate Selection Process 15 Bandwidth estimation Download & measure Pick a rate Initial video rate Video rate for the next video segment Te-Yuan (TY) ACM IMC'12, Boston

TCP Throughput of the Video Flow Playout Buffer is Full 16 TCP Throughput OFF Period: 1-2 seconds Te-Yuan (TY) ACM IMC'12, Boston

Impact of OFF period on TCP TCP sender resets its congestion window – When idle more than one RTO (200ms) – Slow-start restart, RFC 2581/5681 – Linux 3.x (tcp_output.c, line 163) Throughput will be affected – Worse with a competing flow Experience packet loss during slow start 17 50% of the segments get < 1.8Mb/s (Fair Share is 2.5Mb/s) Te-Yuan (TY) ACM IMC'12, Boston

The Rate Selection Process 18 Bandwidth estimation Download & measure Pick a rate Underestimation If perceived 1.8 Mb/s, which video rate would it pick? 1400 kb/s?1750 kb/s? Even lower? Video rate for the next video segment Initial video rate Te-Yuan (TY) ACM IMC'12, Boston

Conservative Rate Selection Available BW Video Rate 19 If 1.8Mb/s BW is perceived, the client switches down to 1050kb/s video rate 19Te-Yuan (TY) ACM IMC'12, Boston

The Rate Selection Process 20 Bandwidth estimation Download & measure Pick a rate Underestimation Conservatively Is there any consequence of being conservative? Initial video rate Video rate for the next video segment Te-Yuan (TY) ACM IMC'12, Boston

Smaller Segment Size for Lower Video Rate 21 When requesting a smaller segment size, lower probability of obtaining fair share. Te-Yuan (TY) ACM IMC'12, Boston

Lower video rate leads to further bandwidth underestimation 22 Lowest Video Rate Highest Video Rate Te-Yuan (TY) ACM IMC'12, Boston

The Rate Selection Process 23 Bandwidth estimation Download & measure Pick a rate Underestimation Conservatively Request for a smaller segment Further Video rate for the next video segment Initial video rate Te-Yuan (TY) ACM IMC'12, Boston

The Complete Story Video Quality 24 Bandwidth Underestimation Conservatism Further Underestimation Te-Yuan (TY) ACM IMC'12, Boston

Conclusion The problem: Video flow gets a much lower throughput than its fair share – Problem occurred in all the services we measured The reason: Video client tries to do TCP’s job – It is afraid of causing network congestion It becomes timid – It tries to figure out its available bandwidth Without seeing the interaction between the playout buffer, segment sizes and TCP dynamics It is easily confused and becomes unstable One service provider changed its algorithm because of this work 25Te-Yuan (TY) ACM IMC'12, Boston