The Future of Transport Hari Balakrishnan LCS and EECS Massachusetts Institute of Technology

Slides:



Advertisements
Similar presentations
A Comparison of Mechanisms for Improving TCP Performance over Wireless Links Published In IEEE/ACM TRANSACTIONS ON NETWORKING, VOL.5 NO.6,DECEMBER 1997.
Advertisements

1 Improving TCP Performance over Mobile Networks HALA ELAARAG Stetson University Speaker : Aron ACM Computing Surveys 2002.
Improving TCP over Wireless by Selectively Protecting Packet Transmissions Carla F. Chiasserini Michele Garetto Michela Meo Dipartimento di Elettronica.
Hui Zhang, Fall Computer Networking TCP Enhancements.
TCP Congestion Control Dina Katabi & Sam Madden nms.csail.mit.edu/~dina 6.033, Spring 2014.
Performance Improvement of TCP in Wireless Cellular Network Based on Acknowledgement Control Osaka University Masahiro Miyoshi, Masashi Sugano, Masayuki.
PERSISTENT DROPPING: An Efficient Control of Traffic Aggregates Hani JamjoomKang G. Shin Electrical Engineering & Computer Science UNIVERSITY OF MICHIGAN,
Prentice HallHigh Performance TCP/IP Networking, Hassan-Jain Chapter 10 TCP/IP Performance over Asymmetric Networks.
Shivkumar Kalyanaraman Rensselaer Polytechnic Institute 1 ECSE-6600: Internet Protocols Informal Quiz #07 Shivkumar Kalyanaraman: GOOGLE: “Shiv RPI”
Transport Layer 3-1 outline r TCP m segment structure m reliable data transfer m flow control m congestion control.
Transport Layer 3-1 Fast Retransmit r time-out period often relatively long: m long delay before resending lost packet r detect lost segments via duplicate.
Transport Layer3-1 Congestion Control. Transport Layer3-2 Principles of Congestion Control Congestion: r informally: “too many sources sending too much.
Open Issues on TCP for Mobile Computing Ibrahim Matta Computer Science, Boston University Vassilis Tsaoussidis Computer Science, Northeastern University.
CS 268: Wireless Transport Protocols Kevin Lai Feb 13, 2002.
1 TCP Transport Control Protocol Reliable In-order delivery Flow control Responds to congestion “Nice” Protocol.
1 Design study for multimedia transport protocol in heterogeneous networks Haitao Wu; Qian Zhang; Wenwu Zhu; Communications, ICC '03. IEEE International.
Transport Layer Congestion control. Transport Layer 3-2 Approaches towards congestion control End-to-end congestion control: r no explicit feedback.
Computer Networks Transport Layer. Topics F Introduction  F Connection Issues F TCP.
TCP in Heterogeneous Network Md. Ehtesamul Haque # P.
Reliable Transport Layers in Wireless Networks Mark Perillo Electrical and Computer Engineering.
CMPE 257 Spring CMPE 257: Wireless and Mobile Networking Spring 2005 E2E Protocols (point-to-point)
L13: Sharing in network systems Dina Katabi Spring Some slides are from lectures by Nick Mckeown, Ion Stoica, Frans.
TCP Congestion Control
1 A Comparison of Mechanisms for Improving TCP Performance over Wireless Links Course : CS898T Instructor : Dr.Chang - Swapna Sunkara.
2: Application Layer 1 1DT066 Distributed Information System Chapter 3 Transport Layer.
Congestion Manager and its relevance to WebTP Rajarshi Gupta.
Wireless TCP February 22, 2002 © 2002 Yongguang Zhang CS 395T - Mobile Computing and Wireless Networks Department of Computer SciencesTHE UNIVERSITY OF.
All rights reserved © 2006, Alcatel Accelerating TCP Traffic on Broadband Access Networks  Ing-Jyh Tsang 
How to get good TCP performance over asymmetric networks Hari Balakrishnan MIT Lab for Computer Science With:
The Internet Congestion Manager Hari Balakrishnan MIT LCS
TCP: flow and congestion control. Flow Control Flow Control is a technique for speed-matching of transmitter and receiver. Flow control ensures that a.
CS640: Introduction to Computer Networks Aditya Akella Lecture 22 - Wireless Networking.
Spring 2000Nitin BahadurAdvanced Computer Networks A Comparison of Mechanisms for Improving TCP Performance over Wireless Links By: Hari B., Venkata P.
Transport Layer 4 2: Transport Layer 4.
Adaptive Transmission Protocols for the Future Internet Hari Balakrishnan MIT Lab for Computer Science
Improving TCP Performance over Mobile Networks Zahra Imanimehr Rahele Salari.
An Integrated Congestion Management Architecture for Internet Hosts Hari Balakrishnan MIT Lab for Computer Science
Principles of Congestion Control Congestion: informally: “too many sources sending too much data too fast for network to handle” different from flow control!
TCOM 509 – Internet Protocols (TCP/IP) Lecture 04_b Transport Protocols - TCP Instructor: Dr. Li-Chuan Chen Date: 09/22/2003 Based in part upon slides.
Wireless TCP Prasun Dewan Department of Computer Science University of North Carolina
Transport Layer Natawut Nupairoj, Ph.D. Department of Computer Engineering Chulalongkorn University.
U Innsbruck Informatik - 1 CADPC/PTP in a nutshell Michael Welzl
Transport over Wireless Networks Myungchul Kim
Rate Adaptation Protocol for Real-time Streams Goal: develop an end-to-end TCP-friendly RAP for semi-reliable rate-based applications (e.g. playback of.
TCP behavior of a Busy Internet Server: Analysis and Improvements Y2K Oct.10 Joo Young Hwang Computer Engineering Research Laboratory KAIST. EECS.
Wireless TCP. References r Hari Balakrishnan, Venkat Padmanabhan, Srinivasan Seshan and Randy H. Katz, " A Comparison of Mechanisms for Improving TCP.
Transport Layer 3-1 Chapter 3 Transport Layer Computer Networking: A Top Down Approach 6 th edition Jim Kurose, Keith Ross Addison-Wesley March
Improving TCP Performance over Wireless Networks
1 Transport Layer Lecture 10 Imran Ahmed University of Management & Technology.
Challenges to Reliable Data Transport Over Heterogeneous Wireless Networks.
Outline Wireless introduction Wireless cellular (GSM, CDMA, UMTS) Wireless LANs, MAC layer Wireless Ad hoc networks – routing: proactive routing, on-demand.
1 Sonia FahmyPurdue University TCP Congestion Control Sonia Fahmy Department of Computer Sciences Purdue University
Thoughts on the Evolution of TCP in the Internet (version 2) Sally Floyd ICIR Wednesday Lunch March 17,
Ασύρματες και Κινητές Επικοινωνίες Ενότητα # 11: Mobile Transport Layer Διδάσκων: Βασίλειος Σύρης Τμήμα: Πληροφορικής.
TCP continued. Discussion – TCP Throughput TCP will most likely generate the saw tooth type of traffic. – A rough estimate is that the congestion window.
An End-System Architecture for Unified Congestion Management Hariharan S. Rahul, Hari Balakrishnan, Srinivasan Seshan MIT Lab for Computer Science
CIS679: UDP and Multimedia r Review of last lecture r UDP and multimedia.
1 Flow & Congestion Control Some slides are from lectures by Nick Mckeown, Ion Stoica, Frans Kaashoek, Hari Balakrishnan, and Sam Madden Prof. Dina Katabi.
Group 5 ECE 4605 Neha Jain Shashwat Yadav
Reddy Mainampati Udit Parikh Alex Kardomateas
COMP 431 Internet Services & Protocols
UNIT-V Transport Layer protocols for Ad Hoc Wireless Networks
The Future of Transport
IT351: Mobile & Wireless Computing
Jiyong Park Seoul National University, Korea
An Integrated Congestion Management Architecture for Internet Hosts
TCP Congestion Control
TCP flow and congestion control
Computer Networks Protocols
Impact of transmission errors on TCP performance
Presentation transcript:

The Future of Transport Hari Balakrishnan LCS and EECS Massachusetts Institute of Technology

Focus Congestion management o New applications o New application traffic patterns Heterogeneous technologies o Wireless o Asymmetric networks o Large and small pipe size technologies

State-of-the-World, Yesterday (& Today!) r1 r-n r3 r2 Independent TCP streams 1. Far too inefficient (multiple slow starts, etc.) 2. More alarmingly, far too aggressive n connections, 1 sees loss; window decreases only by (1 - 1/2n)

State-of-the-World, Today r1 r-n r3 r2 Put everyone on same ordered byte stream While this fixes some of the problems of independent connections, it really is a step in the wrong direction! 1. Far too much coupling between objects 2. Far too application-specific

What is the World Heading Toward? u1 u-m u3 u2 The world won’t be just HTTP The world won’t be just TCP r1 r2 r3 r-n Logically different streams (objects) should be kept separate, yet congestion management must be performed.

What We Really Need… An integrated approach to end-to-end congestion management for the Internet IP Apps Transport instances Congestion management Per-host & per-domain information

Some Salient Features Shared learning Heterogeneous application support Simple application interfaces to congestion manager Robust and stable network behavior Flexible bandwidth-apportioning using receiver hints First step: Transport-Independent Congestion Control (TICC)

Heterogeneous Technologies Non-congestion losses (“errors”) Asymmetry o Bandwidth o Latency (delay variations) Pipe sizes o Large pipes o Small pipes

Errors + Congestion Some people think that we need to split connections to perform well: This is wrong! Careful design of link-layer and transport- aware link protocols work very well Explicit Loss Notification (ELN) helps sender decouple loss recovery from congestion control

Asymmetry Network and traffic characteristics in one direction affect performance in the other Bandwidth, latency (variability), media- access, loss rate… TCP improvements o ACK filtering (purge “redundant” ACKs) o Sender adaptation (rate-controlled transmissions, byte-based window increases) o ACK reconstruction o ACK congestion control (Padmanabhan98)

Pipe sizes Large pipes are problematic o Timeouts when multiple losses occur o SACK fixes this (plus timestamp, PAWS, etc.) o The rtt-bias unfairness problem remains… o How big an rtt before TCP is unusable? Small pipes are the more pressing problem! o Far too many timeouts 55% of all recovery in one traffic trace of a busy Web server (over 1.6 million connections) o A solution: Newreno + Enhanced Recovery (ER) Follow packet conservation, sending new probe packets upon duplicate ACKs No timeouts unless congestion is “persistent”

Conclusions: Revolution or Evolution? A revolution in congestion management o To accommodate heterogeneous applications o But incremental deployability is critical o And then there’s multicast... An evolutionary approach to changing TCP o But with revolutionary “local” techniques o Changes to end-to-end mechanisms (e.g., elements of rate control)