EE689 Lecture 7 Review of Last Lecture Differentiated Services Networks Performance results.

Slides:



Advertisements
Similar presentations
Transport Layer3-1 TCP AIMD multiplicative decrease: cut CongWin in half after loss event additive increase: increase CongWin by 1 MSS every RTT in the.
Advertisements

 Liang Guo  Ibrahim Matta  Computer Science Department  Boston University  Presented by:  Chris Gianfrancesco and Rick Skowyra.
CS640: Introduction to Computer Networks Aditya Akella Lecture 20 – QoS.
Congestion Control: TCP & DC-TCP Swarun Kumar With Slides From: Prof. Katabi, Alizadeh et al.
Mathematical models of the Internet Frank Kelly Hood Fellowship Public Lecture University of Auckland 3 April 2012.
EE 122: Congestion Control The Sequel October 1, 2003.
Router-assisted congestion control Lecture 8 CS 653, Fall 2010.
Presentation by Joe Szymanski For Upper Layer Protocols May 18, 2015.
Introduction 1 Lecture 14 Transport Layer (Transmission Control Protocol) slides are modified from J. Kurose & K. Ross University of Nevada – Reno Computer.
Advanced Computer Networking Congestion Control for High Bandwidth-Delay Product Environments (XCP Algorithm) 1.
Congestion Control An Overview -Jyothi Guntaka. Congestion  What is congestion ?  The aggregate demand for network resources exceeds the available capacity.
XCP: Congestion Control for High Bandwidth-Delay Product Network Dina Katabi, Mark Handley and Charlie Rohrs Presented by Ao-Jan Su.
School of Information Technologies TCP Congestion Control NETS3303/3603 Week 9.
Adaptive Packet Marking for Maintaining End-to-End Throughput in a Differentiated-Services Internet Wu-Chang Feng, Dilip D.Kandlur, Member, IEEE, Debanjan.
1 EE 627 Lecture 11 Review of Last Lecture UDP & Multimedia TCP & UDP Interaction.
Congestion Control Tanenbaum 5.3, /12/2015Congestion Control (A Loss Based Technique: TCP)2 What? Why? Congestion occurs when –there is no reservation.
Proportional Bandwidth Allocation in DiffServ Networks Usman Raza Chohan
TCP Stability and Resource Allocation: Part I. References The Mathematics of Internet Congestion Control, Birkhauser, The web pages of –Kelly, Vinnicombe,
Metrics for Performance Evaluation Nelson Fonseca State University of Campinas.
High speed TCP’s. Why high-speed TCP? Suppose that the bottleneck bandwidth is 10Gbps and RTT = 200ms. Bandwidth delay product is packets (1500.
EE689 Lecture 5 Review of last lecture More on HPF RED.
1 EE 689 Lecture 3 Review of Last Lecture UDP & Multimedia TCP & UDP Interaction.
1 689 Lecture 2 Review of Last Lecture Networking basics TCP/UDP review.
TCP Congestion Control TCP sources change the sending rate by modifying the window size: Window = min {Advertised window, Congestion Window} In other words,
1 EE689 Lecture 4 Review of Last Lecture UDP-based delivery TCP-modifications.
1 TCP Transport Control Protocol Reliable In-order delivery Flow control Responds to congestion “Nice” Protocol.
1 Internet Networking Spring 2003 Tutorial 11 Explicit Congestion Notification (RFC 3168)
A Real-Time Video Multicast Architecture for Assured Forwarding Services Ashraf Matrawy, Ioannis Lambadaris IEEE TRANSACTIONS ON MULTIMEDIA, AUGUST 2005.
Promoting the Use of End-to- End Congestion Control in the Internet Sally Floyd and Kevin Fall Presented by Scott McLaren.
SACRIO - An Active Buffer Mangement Scheme for Differentiaed Services Networks Saikrishnan Gopalakrishnan Cisco Systems Narasimha Reddy Texas A & M University.
1 Manpreet Singh, Prashant Pradhan* and Paul Francis * MPAT: Aggregate TCP Congestion Management as a Building Block for Internet QoS.
TCP in Heterogeneous Network Md. Ehtesamul Haque # P.
Congestion Control for High Bandwidth-delay Product Networks Dina Katabi, Mark Handley, Charlie Rohrs.
CS 268: Lecture 11 (Differentiated Services) Ion Stoica March 6, 2001.
Reducing Flow-Completion Time for Small Flows by Modifying Slow-Start Affan Rauf ( )
Congestion Control for High Bandwidth-Delay Product Environments Dina Katabi Mark Handley Charlie Rohrs.
Introduction 1 Lecture 14 Transport Layer (Congestion Control) slides are modified from J. Kurose & K. Ross University of Nevada – Reno Computer Science.
Transport Layer 4 2: Transport Layer 4.
Adaptive Packet Marking for Providing Differentiated Services in the Internet Wu-chang Feng, Debanjan Saha, Dilip Kandlur, Kang Shin October 13, 1998.
QOS مظفر بگ محمدی دانشگاه ایلام. 2 Why a New Service Model? Best effort clearly insufficient –Some applications need more assurances from the network.
Class-based QoS  Internet QoS model requires per session state at each router  1000s s of flows  per session RSVP is complex => reluctance.
TCP CS 168 Discussion Week 6 Many thanks to past EE 122 GSIs.
Beyond Best-Effort Service Advanced Multimedia University of Palestine University of Palestine Eng. Wisam Zaqoot Eng. Wisam Zaqoot November 2010 November.
CSE679: Prioritized Delivery in UDP and TCP r Prioritized Delivery in UDP r Prioritized Delivery in TCP.
High-speed TCP  FAST TCP: motivation, architecture, algorithms, performance (by Cheng Jin, David X. Wei and Steven H. Low)  Modifying TCP's Congestion.
EE 122: Congestion Control and Avoidance Kevin Lai October 23, 2002.
Congestion Control for High Bandwidth-Delay Product Networks D. Katabi (MIT), M. Handley (UCL), C. Rohrs (MIT) – SIGCOMM’02 Presented by Cheng.
TCP with Variance Control for Multihop IEEE Wireless Networks Jiwei Chen, Mario Gerla, Yeng-zhong Lee.
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.
Thoughts on the Evolution of TCP in the Internet (version 2) Sally Floyd ICIR Wednesday Lunch March 17,
The Macroscopic behavior of the TCP Congestion Avoidance Algorithm.
Explicit Allocation of Best-Effort Service Goal: Allocate different rates to different users during congestion Can charge different prices to different.
Random Early Detection (RED) Router notifies source before congestion happens - just drop the packet (TCP will timeout and adjust its window) - could make.
Flow Control in Multimedia Communication Multimedia Systems and Standards S2 IF Telkom University.
XCP: eXplicit Control Protocol Dina Katabi MIT Lab for Computer Science
NOSSDAV '97 Understanding TCP Dynamics in an Integrated Services Internet Wu-chang Feng, Dilip Kandlur, Debanjan Saha, and Kang Shin.
Increasing TCP's CWND based on Throughput draft-you-iccrg-throughput-based-cwnd-increasing-00 Jianjie You IETF92 Dallas.
An End-to-End Service Architecture r Provide assured service, premium service, and best effort service (RFC 2638) Assured service: provide reliable service.
Univ. of TehranIntroduction to Computer Network1 An Introduction Computer Networks An Introduction to Computer Networks University of Tehran Dept. of EE.
Achievable Service Differentiation with Token Bucket Marking for TCP S. Sahu, D.Towsley University of Massachusetts P. Nain INRIA C. Diot Sprint Labs V.
Quality of Service Frameworks Hamed Khanmirza Principles of Network University of Tehran.
@Yuan Xue A special acknowledge goes to J.F Kurose and K.W. Ross Some of the slides used in this lecture are adapted from their.
@Yuan Xue A special acknowledge goes to J.F Kurose and K.W. Ross Some of the slides used in this lecture are adapted from their.
Corelite Architecture: Achieving Rated Weight Fairness
Topics discussed in this section:
Columbia University in the city of New York
CIS679: Prioritized Delivery in UDP and TCP
EE 122: Lecture 18 (Differentiated Services)
EE 122: Differentiated Services
EE 122: Lecture 10 (Congestion Control)
Presentation transcript:

EE689 Lecture 7 Review of Last Lecture Differentiated Services Networks Performance results

Differentiated Services Keep the core/backbone router stateless Maintain per-flow state at the edges. Mark flows (individual/aggregate) at the edges based on contract profiles. Give better treatment to IN packets and drop OUT packets first at the time of congestion.

2-BIT Proposal Mark packets with Assure/Premium Service tags. Premium Service packets get low delay Assured Service packets get better BW than non-marked packets P packets have higher priority, dropped when out of profile.

Performance Results End-to-end response to packet drops impacts delivered throughput. Flows with same reservations, but different RTTs get different BWs Possible for a flow with no reservation, but better RTT to get more than another flow with reservations.

Performance Results CBR flows can still claim substantial BW - can’t hog all the BW though. Reservations lead to “better best-effort” Service differentiation based on reservations difficult Sharing excess BW difficult - results in loss of service differentiation.

TCP’s Impact Flows with higher target rates fall short and flows with lower rates exceed targets. Multiplicative decrease/additive increase causes this RTT bias inherent in TCP Need better mechanisms for service differentiation

No-OUT OUT packets are more susceptible to be dropped. Every packet drop results in a window decrease Reduce window decreases for higher BW => Don’t send OUT packets?

Inverse-rate drop Drop packets inverse proportion to the reservation Mark packets with reservation level - can keep core stateless Different flows will have packets dropped at different rates Favors higher rate flows

Two-Windows Modify TCP windowing mechanism cwnd = rwnd + ewnd, rwnd = reservation window, ewnd = excess window OUT packet drop ewnd = ewnd/2 IN packet drop cwnd = cwnd/2 Source needs to know how packets are marked - may not work with aggregation

Two-windows Protects reservation level on OUT packet drops More aggressive than current TCP Requires modifications to all end sources.

THREE-DROPS Mark packets IN, OUT-IN, OUT-OUT Maintain state of sending rates of flows at the edge. Mark OUT packets of flows not reaching targets OUT-IN, mark OUT packets of flows exceeding targets OUT-OUT Drop OUT-OUT before OUT-IN Feedback based

Summary Packet Marking and differentiated packet dropping can provide service differentiation Reservation level may not directly correlate to service Improved techniques provide better service differentiation Edge markers can provide substantial improvements