Key insight.  With drop-when-decoded, the busy period of the virtual queue contributes to the physical queue size calculation  Responding to ACK of the.

Slides:



Advertisements
Similar presentations
ROUTING IN INTERMITTENTLY CONNECTED MOBILE AD HOC NETWORKS AND DELAY TOLERANT NETWORKS: OVERVIEW AND CHALLENGES ZHENSHENG ZHANG.
Advertisements

An Easy-to-Decode Network Coding Scheme for Wireless Broadcasting
Congestion Control Reasons: - too many packets in the network and not enough buffer space S = rate at which packets are generated R = rate at which receivers.
24-1 Chapter 24. Congestion Control and Quality of Service (part 1) 23.1 Data Traffic 23.2 Congestion 23.3 Congestion Control 23.4 Two Examples.
Network coding techniques Elena Fasolo Network coding techniques Elena Fasolo PhD Student - SIGNET Group Wireless Systems - Lecture.
Improving TCP Performance over Mobile Ad Hoc Networks by Exploiting Cross- Layer Information Awareness Xin Yu Department Of Computer Science New York University,
Congestion Control Created by M Bateman, A Ruddle & C Allison As part of the TCP View project.
EE 685 presentation Optimal Control of Wireless Networks with Finite Buffers By Long Bao Le, Eytan Modiano and Ness B. Shroff.
CS 408 Computer Networks Congestion Control (from Chapter 05)
On Modeling Feedback Congestion Control Mechanism of TCP using Fluid Flow Approximation and Queuing Theory  Hisamatu Hiroyuki Department of Infomatics.
Networks: Sample Performance Problems 1 Sample Network Performance Problems.
1 Cooperative Communications in Networks: Random coding for wireless multicast Brooke Shrader and Anthony Ephremides University of Maryland October, 2008.
Network Coding and Reliable Communications Group DAWN PI meeting – October 2009 Network coding with unreliable, unknown bandwidths Muriel Medard EECS RLE.
Network Coding Meets TCP Michael Mitzenmacher Joint work with Jay-Kumar Sudararajan, Devavrat Shah, Muriel Medard, Joao Barros.
Protocols. Basics Defining Interactions VERTICAL Application Presentation Session Transport Network Data Link Physical Please do this for me OK It’s.
NCKU CSIE CIAL1 Principles and Protocols for Power Control in Wireless Ad Hoc Networks Authors: Vikas Kawadia and P. R. Kumar Publisher: IEEE JOURNAL ON.
Leveraging Multiple Network Interfaces for Improved TCP Throughput Sridhar Machiraju SAHARA Retreat, June 10-12, 2002.
Little’s Theorem Examples Courtesy of: Dr. Abdul Waheed (previous instructor at COE)
Estimating Congestion in TCP Traffic Stephan Bohacek and Boris Rozovskii University of Southern California Objective: Develop stochastic model of TCP Necessary.
Random coding for wireless multicast Brooke Shrader and Anthony Ephremides University of Maryland Joint work with Randy Cogill, University of Virginia.
1 A State Feedback Control Approach to Stabilizing Queues for ECN- Enabled TCP Connections Yuan Gao and Jennifer Hou IEEE INFOCOM 2003, San Francisco,
Error Checking continued. Network Layers in Action Each layer in the OSI Model will add header information that pertains to that specific protocol. On.
Switching Techniques Student: Blidaru Catalina Elena.
RTS/CTS-Induced Congestion in Ad Hoc Wireless LANs Saikat Ray, Jeffrey B. Carruthers, and David Starobinski Department of Electrical and Computer Engineering.
Integrated Services (RFC 1633) r Architecture for providing QoS guarantees to individual application sessions r Call setup: a session requiring QoS guarantees.
Efficient Network-Coding-Based Opportunistic Routing Through Cumulative Coded Acknowledgments Dimitrios Koutsonikolas, Chih-Chun Wang and Y. Charlie Hu.
When rate of interferer’s codebook small Does not place burden for destination to decode interference When rate of interferer’s codebook large Treating.
Sami Al-wakeel 1 Data Transmission and Computer Networks The Switching Networks.
MARCH : A Medium Access Control Protocol For Multihop Wireless Ad Hoc Networks 성 백 동
User Cooperation via Rateless Coding Mahyar Shirvanimoghaddam, Yonghui Li, and Branka Vucetic The University of Sydney, Australia IEEE GLOBECOM 2012 &
Congestion Control in CSMA-Based Networks with Inconsistent Channel State V. Gambiroza and E. Knightly Rice Networks Group
CprE 545 project proposal Long.  Introduction  Random linear code  LT-code  Application  Future work.
Multicast and Unicast Real-Time Video Streaming Over Wireless LANS April. 27 th, 2005 Presented by, Kang Eui Lee.
Copyright 2008 Kenneth M. Chipps Ph.D. Controlling Flow Last Update
Novel network coding strategy for TDD Use of feedback (ACK) improves delay/energy/ throughput performance, especially for high latency- high errors scenarios.
Wireless TCP. References r Hari Balakrishnan, Venkat Padmanabhan, Srinivasan Seshan and Randy H. Katz, " A Comparison of Mechanisms for Improving TCP.
Deadline-based Resource Management for Information- Centric Networks Somaya Arianfar, Pasi Sarolahti, Jörg Ott Aalto University, Department of Communications.
1 Analysis of a window-based flow control mechanism based on TCP Vegas in heterogeneous network environment Hiroyuki Ohsaki Cybermedia Center, Osaka University,
Nour KADI, Khaldoun Al AGHA 21 st Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications 1.
Internet and Intranet Protocols and Applications The Internet: Delay, Errors, Detection February, 2002 Joseph Conron Computer Science Department New York.
TCP continued. Discussion – TCP Throughput TCP will most likely generate the saw tooth type of traffic. – A rough estimate is that the congestion window.
Network Coding and Reliable Communications Group Modeling Network Coded TCP Throughput: A Simple Model and its Validation MinJi Kim*, Muriel Médard*, João.
Trading Structure for Randomness in Wireless Opportunistic Routing Szymon Chachulski, Michael Jennings, Sachin Katti and Dina Katabi MIT CSAIL SIGCOMM.
Information Theory for Mobile Ad-Hoc Networks (ITMANET): The FLoWS Project Collision Helps! Algebraic Collision Recovery for Wireless Erasure Networks.
Spring Computer Networks1 Congestion Control Sections 6.1 – 6.4 Outline Preliminaries Queuing Discipline Reacting to Congestion Avoiding Congestion.
© Janice Regan, CMPT 128, CMPT 371 Data Communications and Networking Congestion Control 0.
Distance Vector Routing
1)Effect of Network Coding in Graphs Undirecting the edges is roughly as strong as allowing network coding simplicity is the main benefit 2)Effect of Network.
Transmission Control Protocol (TCP) TCP Flow Control and Congestion Control CS 60008: Internet Architecture and Protocols Department of CSE, IIT Kharagpur.
Network Topology Single-level Diversity Coding System (DCS) An information source is encoded by a number of encoders. There are a number of decoders, each.
CS Spring 2011 CS 414 – Multimedia Systems Design Lecture 17 – Multimedia Transport Subsystem (Part 3) Klara Nahrstedt Spring 2011.
Coding for Multipath TCP: Opportunities and Challenges Øyvind Ytrehus University of Bergen and Simula Res. Lab. NNUW-2, August 29, 2014.
Other Methods of Dealing with Congestion
TCP+Network Coding Li Shushi
Topics discussed in this section:
Ad-hoc Networks.
Switching Techniques In large networks there might be multiple paths linking sender and receiver. Information may be switched as it travels through various.
Lecture on Data Link Control
Other Methods of Dealing with Congestion
Congestion Control (from Chapter 05)
Switching Techniques.
Congestion Control (from Chapter 05)
CS4470 Computer Networking Protocols
Congestion Control Reasons:
Congestion Control (from Chapter 05)
Congestion Control (from Chapter 05)
Congestion Control (from Chapter 05)
Congestion Control (from Chapter 05)
Information Theory for Mobile Ad-Hoc Networks (ITMANET): The FLoWS Project Collision Helps! Algebraic Collision Recovery for Wireless Erasure Networks.
ACHIEVEMENT DESCRIPTION
Presentation transcript:

Key insight.  With drop-when-decoded, the busy period of the virtual queue contributes to the physical queue size calculation  Responding to ACK of the degrees of freedom ensures only queuing delay of virtual queues contributes to physical queue size Consequences.  Queue size now grows linearly with 1/(1- ρ)  Reduces the amount of storage needed at intermediate nodes for performing re-encoding  Analysis also applies when only some nodes do re-encoding  ACK of degrees of freedom allows traditional queuing results to be applied easily in scenarios with network coding Queuing analysis for coded networks with feedback J. Sundararajan, D. Shah, M. Médard, M. Mitzenmacher, J. Barros The proposed approach to queue management will play a key role in interfacing TCP with network coding, especially when intermediate nodes re-encode MAIN ACHIEVEMENT:  Propose novel ACK mechanism that allows nodes to manage queue occupancy effectively  Characterize expected queue size at each node Packets can be dropped from queue only upon confirmation of decoding This means the queue sizes will be unnecessarily long In particular, as load factor ρ approaches capacity, queue grows quadratically as a function of 1/(1- ρ) Extend queue management protocol to more general (wireless) scenarios Multipath routing with coding Multicast traffic pattern IMPACT NEXT-PHASE GOALS ACHIEVEMENT DESCRIPTION STATUS QUO NEW INSIGHTS λ x (Time between node ( i -1) seeing pkt and node i seeing pkt) λ x (Time for receiver’s ACK to propagate from source to node k) p 1 p 2 p 3 p 4 p 5 p 6 p 7 p 8 Decoded Seen Unseen Coefficient vectors of received linear combinations, after Gaussian elimination Number of seen packets = Rank of matrix HOW IT WORKS: Acknowledge “seen” packets Almost as if there is link-by-link feedback… ASSUMPTIONS AND LIMITATIONS:  Perfect and delay-free feedback used in analysis, though not critical for the approach  Field size assumed to be very large 1 kN 2 Tx Rx

Problem setup Tandem network of erasure links Bernoulli arrival process of rate λ Perfect delay-free end-to-end feedback (End-to-end nature is motivated by TCP ACKs) Want to study the expected size of the queues at all the nodes N 1k 2

Questions addressed With link-by-link feedback (benchmark): –Every link performs simple ARQ – no coding –Every queue behaves like a Geom/Geom/1 queue –Growth of the queue size as load factor ρ→1 is linear in 1/(1-ρ) With end-to-end feedback: –Need to use intermediate node re-encoding to get to capacity –Degree-of-freedom queue (also called virtual queue) still behaves like a Geom/Geom/1 queue –Can we ensure O(1/(1-ρ)) growth of physical queues in this setting? N 1k 2

Questions addressed Baseline approach: ACK when decoded –Physical queue size is related to busy period of virtual queues –This gives O(1/(1-ρ) 2 ) growth of queues –Also, this approach causes the delay for decoding at the receiver to enter the round-trip time –This has adverse effects in congestion control – TCP windows will close unnecessarily Need to ACK every degree of freedom –Then physical queue size will be related to the waiting time for successful transmission –Then we can achieve O(1/(1-ρ)) growth of queues –TCP window will also progress smoothly, since every incoming packet will generate an ACK without waiting for decoding –How to do this in a way that is simple to implement?

‘Seeing’ a packet p 1 p 2 p 3 p 4 p 5 p 6 p 7 p 8 Decoded SeenUnseen Coefficient vectors of received linear combinations, after Gaussian elimination Number of seen packets = Rank of matrix = Dim of knowledge space Witness for p 4

A new kind of ACK Acknowledge degrees of freedom –ACK a packet upon “seeing” it –Allows ACK of every innovative linear combination, even if it does not reveal a packet immediately p 1 p 2 p 3 p 4 p 5 p 6 p 7 p 8 Decoded SeenUnseen Coefficient vectors of received linear combinations, after Gaussian elimination Witness for p 4

The queue update rule Store every incoming innovative † linear combination Perform row reduction of the stored coefficient matrix and update the packets correspondingly Essentially, queue stores witnesses of seen packets Drop the witness of a packet if you know receiver has seen the packet Implicit ACK: Although only sender gets receiver’s ACK, other nodes can infer receiver’s state from the sender’s coding window, which is embedded in the header † Innovative means the packet is linearly independent of previously received linear combinations

The analysis Use Little’s law to find the expected queue size using expected time spent in queue –Arrival: Packet arrives into queue of node k when the node first sees the packet –Departure: Packet departs when node k finds out that the receiver has seen the packet This duration can be broken into two parts: –T 1 : Time until receiver sees packet –T 2 : Time till node k learns of receiver’s ACK Lemma: Let S A and S B be the set of packets seen by two nodes A and B respectively. Assume S A \S B is non-empty. Suppose A sends a random linear combination of its witnesses of packets in S A and B receives it successfully. The probability that this transmission causes B to see the oldest packet in S A \S B is (1 − 1/q), where q is the field size.

The analysis (contd.) Lemma implies that the virtual queues behave like a FIFO Geom/Geom/1 queue Hence, the time between node i seeing a packet and node i+1 seeing the packet is the waiting time in a Geom/Geom/1 queue, with expectation: Hence, time till receiver sees packet is: Additional time till receiver’s ACK propagates to node k is Hence, using Little’s law, the expected queue size is:

Conclusions Proposed a new ACK mechanism that acknowledges every degree of freedom Analyzed expected queue length for single path with re- encoding at one or more intermediate nodes, and end-to- end feedback Queue size now grows linearly with 1/(1- ρ) Need to extend the protocol and analysis to the case of multiple paths and multiple receivers