Advanced Image Information (3) 画像情報特論 (3) TCP Variants. 情報理工学専攻甲藤二郎
|
|
- Irene Montgomery
- 5 years ago
- Views:
Transcription
1 画像情報特論 (3) Advanced Image Information (3) TCP Variants 情報理工学専攻甲藤二郎
2 TCP Variants
3 TCP-Reno (loss-based) cwnd loss loss loss a= buffer b=0.5 0 BDP increase: cwnd = cwnd + /cwnd decrease: cwnd = cwnd / n AIMD: additive increase multiplicative decrease
4 TCP-Vegas (delay-based) cwnd buffer stored packets in a buffer α BDP n 0 e.g. α=, β=3 diff cwnd RTT min cwnd RTT RTT min stored packets in a buffer increase: cwnd cwnd cwnd cwnd diff otherwise diff decrease: cwnd cwnd *0.75
5 TCP problems, 5 years ago broadband wired networks slow window increase (inefficiency) deployment of wireless networks cannot distinguish wireless errors and buffer overflow TCP-Reno (NewReno, SACK) problem Reno expels Vegas (unfriendliness)
6 TCP Variants in the th century Loss-driven (AIMD) TCP-Reno / NewReno / SACK High-Speed TCP (IETF RFC 3649, Dec 003) Scalable TCP (PFLDnet 003) BIC-TCP / CUBIC-TCP (IEEE INFOCOM 004, PFLDnet 005)... Linux H-TCP (PFLDnet 004) TCP-Westwood (ACM MOBICOM 00) Delay-driven (RTT Observation) TCP-Vegas (IEEE JSAC, Oct 995) FAST-TCP (INFOCOM 004) Hybrid Gentle High-Speed TCP (PfHSN 003) TCP-Africa (IEEE INFOCOM 005) Compound TCP (PFLDnet 006)... Windows (proposed by MSR) Adaptive Reno (PFLDnet 006) YeAH-TCP (PFLDnet 007) TCP-Fusion (PFLDnet 007) our lab
7 Loss-based TCPs aggressive adaptive a Variants Increase / Update Decrease TCP-Reno 0.5 HighSpeed TCP (HS-TCP) w a( w) b( w) p( w) b( w) e.g. 70 (0Gbps, 00ms) Scalable TCP (STCP) 0.0 (per every ACK) BIC-TCP CUBIC-TCP additive increase ( fast) binary search ( slow) max probing ( fast) w 3 0.4( t 3 W max ) Wmax H-TCP 0. 5 t TH TCP-Westwood (TCPW) b log( w) log( Wlow ) b( w) ( bhigh 0.5) 0.5 log( W ) log( W ) high low e.g. 0. (0Gbps, 00ms) RTT RTT RE * RTT BE * RTT min max min min B( k ) B( k) for B( k) / PS / PS otherwise ( congested) ( not congested)
8 Delay-based TCPs Variants Update Decrease TCP-Vegas ) ( ) ( ) ( congestion early w stable w congestion no w w 0.75 FAST-TCP w RTT RTT w w w min ) (, min 0.5 (?) a b
9 Hybrid TCP cwnd loss loss loss buffer 0 BDP n RTT increase: loss mode improvement of friendliness no RTT increase: delay mode improvement of efficiency
10 Hybrid TCPs a b simple adaptive Variants Increase Decrease Gentle HS-TCP HS-TCP / Reno HS-TCP TCP-Africa HS-TCP / Reno HS-TCP Compound TCP (CTCP) 0.5 cwnd / Reno 0.5 Adaptive Reno (ARENO) YeAH-TCP TCP-Fusion B /0Mbps / Reno STCP / Reno B * D PS min 0.5 max ( non congestion loss) ( congestion loss) RTTmin RTT, 0.5 / Reno RTTmin max, 0.5 RTT
11 CUBIC-TCP (Linux default)
12 BIC-TCP () Binary Increase Congestion Control パケットロスが発生した cwnd 平衡点 L.Xu et al: Binary Increase Congestion Control (BIC) for Fast Long-Distance Networks, IEEE INFOCOM 004.
13 BIC-TCP () Window Increase concave convex additive increase (linear increase) Wmax: cwnd when a last loss happened Smax: maximum increase rate (e.g. 3) Smin: minimum increase rate (e.g. 0.0) if (cwnd < Wmax ) Winc = (Wmax cwnd) / ; else Winc = (cwnd - Wmax) / ; if (Winc > Smax) Winc = Smax; elseif (Winc < Smin) Winc = Smin; binary search cwnd = cwnd + Winc / cwnd; L.Xu et al: Binary Increase Congestion Control (BIC) for Fast Long-Distance Networks, IEEE INFOCOM 004.
14 BIC-TCP (3) Window Decrease Wmax=target cwnd update loss Wmax,new=0.9*cwnd loss Wmax,new=cwnd if (cwnd < Wmax ) Wmax,new = cwnd * (-) / ; else Wmax,new = cwnd; cwnd = cwnd * (- ); : decrease rate (e.g. 0.) *0.9: give bandwidth to newly-coming flows... Fast Convergence L.Xu et al: Binary Increase Congestion Control (BIC) for Fast Long-Distance Networks, IEEE INFOCOM 004.
15 CUBIC-TCP () Cubic approximation of BIC-TCP S.Ha et al: CUBIC: A New TCP Friendly HighSpeed TCP Variant, ACM SIGOPS Review, 008.
16 CUBIC-TCP () Window Increase W(t) /* cubic function */ Winc = W(t+RTT) cwnd; cwnd = cwnd + Winc / cwnd; K Wmax*(-) W ( t) C *( t K) K W tcp 3 W max W 3 max equivalent to Reno C t ( t) Wmax ( ) 3 RTT /* TCP mode */ if ( Wtcp > cwnd ) cwnd = Wtcp; window decrease is the same as BIC : decrease rate (e.g. 0.) C: constant (e.g. 0.4) S.Ha et al: CUBIC: A New TCP Friendly HighSpeed TCP Variant, ACM SIGOPS Review, 008.
17 CUBIC-TCP (3) CUBIC s cwnd behavior S.Ha et al: CUBIC: A New TCP Friendly HighSpeed TCP Variant, ACM SIGOPS Review, 008.
18 CUBIC-TCP (4) Advantages stability intra-protocol fairness among multiple CUBIC flows Disadvantages heavy buffer occupancy and delay increase ( delay-based) inter-protocol unfairness against other TCP flows Linux beats Windows! (vs. Compound TCP) K.Munir et al: Linux beats Windows! or the Worrying Evolution of TCP..., PFLDNet 007.
19 Hybrid TCPs
20 Hybrid TCP () single flow packet loss clearing buffer (TCPW) cwnd last cwnd last RTT RTT min vacant capacity Hybrid cwnd last / legacy (Reno) BDP n adaptive switching of two modes (loss & delay): constant rate until RTT increases (delay mode) : efficiency and low delay performs as Reno when RTT increases (loss mode) : friendliness
21 Hybrid TCP () two flows packet loss clearing buffer (TCPW) cwnd last cwnd last RTT RTT min vacant capacity Hybrid 3 cwnd last / legacy (Reno) BDP/ n adaptive switching of two modes (loss & delay): fast cwnd increase (delay... efficiency ) mild cwnd decrease (delay... congestion avoidance) 3 performs as Reno when RTT increases (loss... friendliness )
22 Min-Max Fair (ideal case) Min-Max-Fair: allocate maximum bandwidth to a user who has minimum bandwidth 4 3 allocate (remaining) to a flow allocate (fair) /3 to each flow allocate remaining /3 to a flow 3 5 D.Bertsekas and R.Gallager: Data Networks, Prentice Hall.
23 Ideal: TCP s objective bandwidth session start Min-Max Fair another session 時間 TCP Reno bandwidth session start searching for Min-Max Fair another session 時間
24 TCP behavior model () model definition Loss-mode (TCP-Reno) : cwnd += (per RTT round ) cwnd *= / (when a packet loss is detected) Delay-mode : fill a pipe (fully utilize a link) without causing RTT increase Hybrid : works in delay mode when RTT is not increased works in loss mode when RTT is increases (i.e. when packets are buffered) mode selection: cwnd = max( cwnd loss, cwnd delay )
25 TCP behavior model () parameter definition w : cwnd when a packet loss is detected W : cwnd which just fills a pipe ~ BDP p : packet loss rate assumption packet loss due to buffer overflow is equivalent to packet loss due to random error p 8 3w (in case of TCP-Reno)
26 TCP behavior model (3) TCP friendly model cwnd TCP-Reno w w/ R w: cwnd when a packet loss is detected p: packet loss rate RTT: round trip time 0 w/ RTT round n R: TCP equivalent rate # of transmitted packets until a packet loss is detected = area of a trapezoid w w w 3w 8 p R 8 3w PS RTT 3 p
27 TCP behavior model (4) single flow sender bottleneck link receiver single TCP flow
28 TCP behavior model (5) cwnd & RTT behaviors of ideal models (single flow case) cwnd w loss & hybrid w/ W 0 delay w/ buffer BDP n hybrid cwnd loss w hybrid W delay w/ 0 W-w/ w/ loss-driven delay-driven vacant capacity buffer BDP n cwnd delay & hybrid W w w/ 0 w ~PLR W~bandwidth vacant capacity loss buffer BDP n w/ RTT loss & hybrid delay RTT min 0 w/ n RTT loss & hybrid RTT min delay 0 W-w/ w/ n RTT min RTT 0 w/ loss, delay & hybrid n (i) W < w/ (ii) w/<w <w (iii) w <W large buffer, small PLR (always loss-mode) small buffer, medium PLR (delay loss) large PLR, always vacant (always delay-mode)
29 TCP behavior model (6) formulation TCP CA round (i) W < w/ (ii) w/ W < w (iii) w W Loss transmitted packets elapsed time 3 w 8 w RTTmin (3w 4wW ) 8 PS B 3 3 w w 8 8 w RTTmin ( w W) PS B w RTT min Delay transmitted packets ww ww w W elapsed time w RTT min w RTT min w RTT min Hybrid transmitted packets 3 w 8 ww ( w W ) w W elapsed time w RTTmin (3w 4wW ) 8 PS B w RTTmin ( w W ) PS B w RTT min PS: Packet size, B: Link bandwidth
30 TCP behavior model (7) abstraction of actual hybrids Hybrids Window increase Window decrease Compound TCP 0.5*cwnd 0.75 / ARENO B/0Mbps /~ YeAH-TCP Scalable TCP (.0) /, RTT min /RTT, 7/8 TCP-Fusion B*D min /(N*PS) RTT min /RTT D min : timer resolution, N: # of flows
31 TCP behavior model (8) evaluation by models and simulations throughput 0 Throughput (Mbps) Case Case Case 3 loss-driven delay & hybrid Reno (sim) FAST (sim) CTCP (sim) ARENO (sim) YeAH (sim) Fusion (sim) Loss (model) Delay (model) Hybrid/Fusion (model) CTCP (model) ARENO (model) YeAH (model) Packet Loss Rate (0^n) loss rate Gbps Gbps 00Mbps RTT=40ms buffer size = BDP (constant) Packet loss rate : variable when PLR is large (w/<w), throughputs of delay & hybrid are much larger than that of loss-mode (i.e. efficiency) degradation of Compound & YeAH is due to fixed window decrease
32 TCP behavior model (9) two flows (competing) loss-based TCP flow bottleneck link senders receivers loss-based or hybrid TCP flow
33 TCP behavior model (0) cwnd behavior of ideal models (two flow case) cwnd loss-driven hybrid total (loss + hybrid) w ~PLR W~bandwidth w W w/ 0 total loss & hybrid w/ buffer BDP n cwnd W w W/ w/ 0 total hybrid loss (W-w)/ w/ buffer BDP n cwnd W W/ w w/ 0 total hybrid loss w/ buffer BDP n (i) W < w (low PLR) (ii) w <W < *w (medium PLR) (iii) *w <W (high PLR) always buffered (loss mode) vacant buffered (delay loss) always vacant (delay mode) large buffer, small PLR small buffer, medium PLR large PLR, always vacant
34 TCP behavior model () formulation TCP CA round (i) W < w (ii) w W < w (iii) w W Loss Hybrid (common) transmitted packets transmitted packets elapsed time 3 w w w ( W w) w RTTmin w(3w W ) 4 PS B 3 3 w w 8 8 w RTTmin (w W) 4 PS B 3 w W w 8 w RTT min PS: Packet size, B: Link bandwidth
35 TCP behavior model () evaluation by models and simulations throughput Throughput (Mbps) Case Case Case 3 delay & hybrid loss-driven Reno (sim) FAST (sim) CTCP (sim) ARENO (sim) YeAH (sim) Fusion (sim) Loss (model) Hybrid (model) CTCP (,model) ARENO (model) YeAH (model) Fusion (model) Gbps Gbps 00Mbps RTT=40ms buffer size = BDP (constant) Packet loss rate : variable when PLR is large (w<w), throughputs of delay & hybrid are much larger than that of loss-mode (efficiency) Packet Loss Rate (0^n) loss rate when PLR is low (w>w), hybrid behaves similar to loss-mode (friendliness)
36 TCP behavior model (3) Advantages of Hybrid TCP when vacant capacity exists (or PLR is large), throughput efficiency is greatly improved (advantage of delay-mode) when no vacant capacity exists (or buffer size is large), friendliness to legacy TCP (i.e. Reno) is achieved (advantage of loss-mode) Disadvantages of Hybrid TCP when buffer size is large, delay-mode is never activated
37 Summary of Hybrid TCP Efficiency, Friendliness and Low delay can be applied to real-time streaming and large file download might be effective in wireless networks friendliness to CUBIC-TCP or Compound-TCP CUBIC-TCP: Linux default Compound-TCP: Windows other metrics RTT fairness, mice/elephant (short-lived or longlived), convergence speed, etc efficiency is brought by delay-mode
38 Summary
39 Summary of TCP versions CUBIC-TCP provides efficiency, but tends to increase latency because router buffers are filled up Compound-TCP provides low delay thanks to its delay mode, but suffers from unfriendliness against CUBIC-TCP Some community discusses redesign of TCP
Copa: Practical Delay-Based Congestion Control
Copa: Practical Delay-Based Congestion Control Venkat Arun and Hari Balakrishnan MIT, CSAIL web.mit.edu/copa The Internet is more challenging than ever (Why are we still talking about congestion control
More informationEnhancements for RTT-Fair HighSpeed TCP
Enhancements for RTT-Fair HighSpeed TCP Damien Phillips, Jiankun Hu School of Computer Science and Information Technology, RMIT University GPO Box 2476V, Melbourne 31, Australia Email: {damphil, jiankun}@cs.rmit.edu.au
More informationBBR Congestion Control
BBR Congestion Control Neal Cardwell, Yuchung Cheng, C. Stephen Gunn, Soheil Hassas Yeganeh, Van Jacobson IETF 97: Seoul, Nov 2016 1 Congestion and bottlenecks 2 Delivery rate Congestion and bottlenecks
More informationTCP-Friendly Equation-Based Congestion Control
TCP-Friendly Equation-Based Congestion Control (widmer@informatik.uni-mannheim.de) 09 December 2002 Overview Introduction to congestion control Equation-based congestion control (TFRC) Congestion control
More information画像情報特論 (2) ハイブリッド TCP 情報理工学専攻甲藤二郎.
画像情報特論 () ハイブリッド TCP 情報理工学専攻甲藤二郎 E-Mail: katto@waseda.jp ネットワーク シミュレーション & エミュレーション ネットワークの研究 アルゴリズム 理論解析 シミュレーションション エミュレーション 実装 シミュレータとエミュレータ (1) シミュレーション エミュレーション ソフトウェア PC / 専用機 0 4 0 4 3 3 1 5 ノード
More informationFairness of Congestion-Based Congestion Control: Experimental Evaluation and Analysis
Fairness of Congestion-Based Congestion Control: Experimental Evaluation and Analysis Shiyao Ma, Jingjie Jiang, Wei Wang, Bo Li Department of Computer Science and Engineering, Hong Kong University of Science
More informationA Proactive Approach to Avoid Performance Degradation of MPTCP
A Proactive Approach to Avoid Performance Degradation of MPTCP Draft-zuo-mptcp-degradation-00.txt Authors: F. Wang, Jing Zuo, Z. Cao, K. Zheng, Huawei 1 MPTCP on high-speed rails Two LTE (4G) on high-speed
More informationPassive TCP Stream Estimation of RTT and Jitter Parameters. Jason But Urs Keller Grenville Armitage David Kennedy
Passive TCP Stream Estimation of RTT and Jitter Parameters Jason But Urs Keller Grenville Armitage David Kennedy Outline Passive Monitoring of TCP RTT and Jitter Estimation algorithm Accuracy of Algorithm
More informationBinary Increase Congestion Control (BIC) for Fast Long-Distance Networks
Binary Increase Congestion Control () for Fast Long-Distance Networks Lisong Xu, Khale Harfoush, an Injong Rhee Department of Computer Science North Carolina State University Raleigh, NC 27695-7534 lxu2,
More informationTIMELY: RTT-based Congestion Control for the Datacenter
TIMELY: RTT-based Congestion Control for the Datacenter Authors: Radhika Mittal(UC Berkeley), Vinh The Lam, Nandita Dukkipati, Emily Blem, Hassan Wassel, Monia Ghobadi(Microsoft), Amin Vahdat, Yaogong
More informationInternet Engineering Task Force (IETF) Category: Standards Track July 2011 ISSN:
Internet Engineering Task Force (IETF) G. Renker Request for Comments: 6323 G. Fairhurst Updates: 4342, 5622 University of Aberdeen Category: Standards Track July 2011 ISSN: 2070-1721 Abstract Sender RTT
More informationTCP IN NETWORKS WITH ABRUPT DELAY VARIATIONS AND RANDOM LOSS
TCP N NETWORKS WTH ABRUPT DELAY VARATONS AND RANDOM LOSS Alhussein A. Abouzeid, Sumit Roy Department of Electrical Engineering University of Washington, Box 352500 Seattle, WA 98195-2500 e-mail:{hussein,roy}@ee.washington.edu
More informationAn Empirical Mixture Model for Large-Scale RTT Measurements
1 An Empirical Mixture Model for Large-Scale RTT Measurements Romain Fontugne 1,2 Johan Mazel 1,2 Kensuke Fukuda 1,3 1 National Institute of Informatics 2 JFLI 3 Sokendai June 9, 2015 Introduction RTT:
More informationRTTometer: Measuring Path Minimum RTT with Confidence
RTTometer: Measuring Path Minimum RTT with Confidence Amgad Zeitoun Zhiheng Wang Sugih Jamin Department of Electrical Engineering and Computer Science, The University of Michigan, Ann Arbor {azeitoun,zhihengw,jamin}@eecs.umich.edu
More information(12) Patent Application Publication (10) Pub. No.: US 2009/ A1
(19) United States US 2009.02451 05A1 (12) Patent Application Publication (10) Pub. No.: US 2009/0245105 A1 HO (43) Pub. Date: Oct. 1, 2009 (54) METHOD FOR NETWORK TRANSMISSION (30) Foreign Application
More informationMAC Sleep Mode Control Considering Downlink Traffic Pattern and Mobility
1 MAC Sleep Mode Control Considering Downlink Traffic Pattern and Mobility Neung-Hyung Lee and Saewoong Bahk School of Electrical Engineering & Computer Science, Seoul National University, Seoul Korea
More informationTIMELY: RTT-based Congestion Control for the Datacenter
TIMELY: RTT-based Congestion Control for the Datacenter Radhika Mittal*(UC Berkeley), Vinh The Lam, Nandita Dukkipati, Emily Blem, Hassan Wassel, Monia Ghobadi*(Microsoft), Amin Vahdat, Yaogong Wang, David
More informationPower Management for Networks to Reduce Energy Consumption
Power Management for Networks to Reduce Energy Consumption David Wetherall Intel Research & University of Washington Joint work with: Sylvia Ratnasamy (Intel Research) Sergiu Nedevschi (UC Berkeley) Lucian
More informationAC Disconnect & more. Roger Karam Las Vegas Interim May 2002
AC Disconnect & more Roger Karam Las Vegas Interim May 2002 1 Testing Summary Tests covered AC Disconnection and DC Detection No Apparent Show Stoppers Biggest Concern is Noise Affecting Ethernet s Integrity
More informationLarge-Scale RTT Measurements from an Operational UMTS/GPRS Network
Large-Scale RTT Measurements from an Operational UMTS/GPRS Network Francesco Vacirca, Fabio Ricciato, Rene Pilz Telecommunications Research Center Vienna (ftw.) Donau-City St., 22 Vienna {vacirca,ricciato,pilz}@ftw.at
More informationFrequently Asked Questions
Frequently Asked Questions How does the SMRT-Y work? The SMRT-Y sensor accurately measures the moisture level in the soil. When the soil is dry, the SMRT-Y allows the irrigation controller to water as
More informationDebloating the Linux WiFi Stack
Debloating the Linux WiFi Stack Toke Høiland-Jørgensen Karlstad University toke.hoiland-jorgensen@kau.se 1 QCA, 9th Dec 2016 Toke Høiland-Jørgensen Outline A history of bloat fixes in Linux The issues
More informationEPON Protocol over Coax
TDD Cycle Steve Shellhammer (Qualcomm) June 2013 1 Description of TDD Cycle EPON Protocol over Coax The TDD cycle consists of four time segments Downstream (DS) Upstream (US) Two Guard Times (GT) DS GT
More informationPassive Online RTT Estimation for Flow-Aware Routers using One-Way Traffic
Passive Online RTT Estimation for Flow-Aware Routers using One-Way Traffic Damiano Carra, 1 Konstantin Avrachenkov, 2 Sara Alouf, 2 Alberto Blanc, 2 Philippe Nain, 2 and Georg Post 3 1 University of Verona,
More informationRelationship between reaction time, mental processing speed and motor speed in individuals with mild to moderate brain injury
Relationship between reaction time, mental processing speed and motor speed in individuals with mild to moderate brain injury Kenneth NK FONG 1, PhD, Marko KL CHAN 2, BSc, Peggie PK NG 2, MSc 1 Department
More informationFREECLIMBER 4400/4600 PT/CL CONSOLE
The StairMaster FreeClimber 4400/4600 PT/CL systems console is divided into seven sections: the display window, the workout options, the numeric keypad, the entertainment keypad, the workout statistics,
More informationTransitioning from TTY and Legacy Text-to-911 to Advanced Real-Time Text (RTT)
1 Transitioning from TTY and Legacy Text-to-911 to Advanced Real-Time Text (RTT) Christian Militeau Director, Technical Standards West Safety Services cmiliteau@west.com 720-864-5245 2 Agenda Real Time
More informationPeering vs. Transit. Adnan Ahmed University of Iowa. University of Iowa
Peering vs. Transit Adnan Ahmed Introduction Transit Provides connectivity to the Internet Traffic volume based fees Peering Bilateral exchange Settlement-free (no fee) 1. Introduction Related work Interconnection
More informationShoreTel Network Services Portfolio FAQ
FREQUENTLY ASKED QUESTIONS ShoreTel Network Services Portfolio FAQ You have questions. We have answers. What is the ShoreTel Network Services Portfolio? Leveraging the proven methodology and best practices
More information5 REASONS TO CHOOSE THE E-LINK 1000EXR OVER COMPETING 70/80 GIGABIT WIRELESS SOLUTIONS
5 REASONS TO CHOOSE THE E-LINK 1000EXR OVER COMPETING 70/80 GIGABIT WIRELESS SOLUTIONS EXECUTIVE SUMMARY 70/80 GHz e-band products are being touted by many global mobile operators as an integral component
More informationPinpointing Delay and Forwarding Anomalies Using Large-Scale Traceroute Measurements
Pinpointing Delay and Forwarding Anomalies Using Large-Scale Traceroute Measurements Romain Fontugne 1, Emile Aben 2, Cristel Pelsser 3, Randy Bush 1 November 1, 2017 1 IIJ Research Lab, 2 RIPE NCC, 3
More informationParticle Therapy Systems by Mitsubishi Electric DG1101-KM-0034
Particle Therapy Systems by Mitsubishi Electric DG1101-KM-0034 Saudi-Japan Business Opportunities Forum February 1 2, 2012 1 2 Types of Treatment Conventional Radiation Therapy Uses photons (energetic
More informationUser Guide for Lifegear RUNNER App
User Guide for Lifegear RUNNER App http://www.lifegear.com.tw Contents LifeGear RUNNER App Installation 2 Connect Your Mobile Device to the Specific Bluetooth Fitness Equipment 4 Entering Your Profile
More informationPinpointing Anomalies in Large-Scale Traceroute Measurements. Romain Fontugne & Kenjiro Cho November 10, 2016
Pinpointing Anomalies in Large-Scale Traceroute Measurements Romain Fontugne & Kenjiro Cho November 10, 2016 1 / 31 This presentation On going research work conducted at IIJ-II In collaboration with: Emile
More informationParametric Optimization and Analysis of Adaptive Equalization Algorithms for Noisy Speech Signals
IOSR Journal of Electrical and Electronics Engineering (IOSR-JEEE) e-issn: 2278-1676, p-issn: 2320 3331, Volume 4, Issue 6 (Mar. -Apr. 2013), PP 69-74 Parametric Optimization and Analysis of Adaptive Equalization
More informationCONSOLE GUIDE & SET-UP INSTRUCTIONS
CONSOLE GUIDE & SET-UP INSTRUCTIONS TABLE OF CONTENTS MTX Console Guide Console Features... 1 Dynamic Display (LCD)... 2 Keypad Buttons... 5 Software Management... 7 Optional Accessories... 8 CROSS CiRCUIT
More informationFog Computing: Challenges and Solutions
Fog Computing: Challenges and Solutions Maria Gorlatova @MariaGorlatova My Background (1/2) Participating in GHC since 2012 Baltimore 12 Phoenix 14 Houston 15, 16 Helping organize GHC since 2015 Co-chair,
More informationJitter-aware time-frequency resource allocation and packing algorithm
Jitter-aware time-frequency resource allocation and packing algorithm The MIT Faculty has made this article openly available. Please share how this access benefits you. Your story matters. Citation As
More informationLACNIC Annual Meeting Monday, May 25, 2009 Bill Woodcock Research Director Packet Clearing House
Introduction to Peering & Internet Exchange Points LACNIC Annual Meeting Monday, May 25, 2009 Bill Woodcock Research Director Packet Clearing House ISP Lifecycle: Simple Aggregator Single Transit Provider
More informationTREADMILL. Vision Fitness Classic Console TF40
TREADMILL Vision Fitness Classic Console TF40 CLASSIC CONSOLE OPERATION A) LARGE LED DISPLAY WINDOW - displays workout time B) ALPHANUMERIC DISPLAY WINDOW displays incline, distance, speed, calories, pace,
More informationCONSOLE GUIDE & SET-UP INSTRUCTIONS STANDARD CONSOLE
CONSOLE GUIDE & SET-UP INSTRUCTIONS STANDARD CONSOLE ii TABLE OF CONTENTS Standard Console Guide ZR7000 / ZR8000 Console Features... 1 Dynamic Display (LCD)... 2 Keypad Buttons... 4 User (Club) Settings...
More informationModern Technology of Internet
Modern Technology of Internet Jiří Navrátil, Josef Vojtěch, Jan Furman, Tomáš Košnar, Sven Ubik, Milan Šárek, Jan Růžička, Martin Pustka, Laban Mwansa, Rudolf Blažek Katedra počítačových systémů FIT České
More informationUser-Friendly Approach to Capacity Planning studies with Java Modelling Tools
Politecnico di Milano EECS Dept. Milan, Italy User-Friendly Approach to Capacity Planning studies with Java Modelling Tools Marco Bertoli, Giuliano Casale, Giuseppe Serazzi SIMUTOOLS09 March 5th, 2009
More informationThe Effect of Extremes in Small Sample Size on Simple Mixed Models: A Comparison of Level-1 and Level-2 Size
INSTITUTE FOR DEFENSE ANALYSES The Effect of Extremes in Small Sample Size on Simple Mixed Models: A Comparison of Level-1 and Level-2 Size Jane Pinelis, Project Leader February 26, 2018 Approved for public
More informationBefore the Federal Communications Commission Washington, D.C
Before the Federal Communications Commission Washington, D.C. 20554 In the Matter of ) ) Transition From ) TTY to Real-Time Text ) CG Docket No. 16-145 Technology ) ) ) ) COMMENTS OF THE ALLIANCE FOR TELECOMMUNICATIONS
More informationOn Studying the Impact of the Internet Delays on Audio Transmission
On Studying the Impact of the Internet Delays on Audio Transmission Lopamudra Roychoudhuri, Ehab Al-Shaer, Hazem Hamed and Gregory B. Brewster Multimedia Networking Research Laboratory School of Computer
More informationCourse Syllabus. Operating Systems, Spring 2016, Meni Adler, Danny Hendler and Amnon Meisels 1 3/14/2016
Course Syllabus. Introduction - History; Views; Concepts; Structure 2. Process Management - Processes; State + Resources; Threads; Unix implementation of Processes 3. Scheduling Paradigms; Unix; Modeling
More informationTTY/TDD Minimum Performance Specification
GPP C.S00-B Version.0 May 0 TTY/TDD Minimum Performance Specification 0 GPP GPP and its Organizational Partners claim copyright in this document and individual Organizational Partners may copyright and
More informationDesign Strategies for ARX with Provable Bounds: SPARX and LAX
Design Strategies for ARX with Provable Bounds: SPARX and LAX Daniel Dinu 1, Léo Perrin 1, Aleksei Udovenko 1, Vesselin Velichkov 1, Johann Großschädl 1, Alex Biryukov 1 1 SnT, University of Luxembourg
More informationCONSOLE GUIDE & SET-UP INSTRUCTIONS STANDARD CONSOLE
Time Calories Distance Pace Length Height HR% HR Level GET STARTED IN 5 SIMPLE STEPS 1 HOP FROM SIDE TO SIDE 2 START WITH SHORT STRIDES 3 KICK YOUR HEELS TO YOUR GLUTES 4 SLOWLY EXTEND YOUR STRIDE 5 MOVE
More informationBimodal Multicast. Harald Gjermundrød CptS/EE 562 February 10, 2004
1 Bimodal Multicast Harald Gjermundrød CptS/EE 562 February 10, 2004 Bimodal Multicast 2 Bimodal Attack General commands soldiers Consensus is impossible! If almost all attack victory is certain If very
More informationCHARACTERIZATION OF ANNULAR ARRAY TRANSDUCER
Analele Universităţii de Vest din Timişoara Vol. LV, 2011 Seria Fizică CHARACTERIZATION OF ANNULAR ARRAY TRANSDUCER Luminita Moraru 1, Laura Onose 1, 2, Ana-Maria Chiselev 1 1 Dunărea de Jos University
More informationCFS for Addressing CPU Resources in Multi-Core Processors with AA Tree
CFS for Addressing CPU Resources in Multi-Core Processors with AA Tree Prajakta Pawar, S.S.Dhotre, Suhas Patil Computer Engineering Department BVDUCOE Pune-43(India) Abstract- CPU scheduler plays crucial
More informationPart I. Boolean modelling exercises
Part I. Boolean modelling exercises. Glucose repression of Icl in yeast In yeast Saccharomyces cerevisiae, expression of enzyme Icl (isocitrate lyase-, involved in the gluconeogenesis pathway) is important
More informationAPPENDIX D STRUCTURE OF RAW DATA FILES
APPENDIX D STRUCTURE OF RAW DATA FILES The CALCAP program generates detailed records of all responses to the reaction time stimuli. Data are stored in a file named subj#-xx.dat. Where subj# is the subject
More informationPrentice Hall. A Survey of Mathematics with Applications, 7th Edition Mississippi Mathematics Framework 2007 Revised,
Prentice Hall A Survey of Mathematics with Applications, 7th Edition 2005 C O R R E L A T E D T O Mississippi Mathematics Framework 2007 Revised, Survey of Mathematical Topics NUMBER AND OPERATIONS Competencies
More informationBefore the FEDERAL COMMUNICATIONS COMMISSION Washington, D.C ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) PETITION FOR RULEMAKING
Before the FEDERAL COMMUNICATIONS COMMISSION Washington, D.C. 20554 In the Matter of Facilitating the Deployment of Text-to-911 and Other Next Generation 911 Applications Framework for Next Generation
More informationRocky K. C. Chang, Edmond Chan, Waiting Fok, and Weichao Li. The Hong Kong Polytechnic University Hung hom, Kowloon, Hong Kong
Rocky K. C. Chang, Edmond Chan, Waiting Fok, and Weichao Li The Hong Kong Polytechnic University Hung hom, Kowloon, Hong Kong APRICOT 2010 1 Problem statement Measurement system Measurement methodology
More informationThe Loss of Heterozygosity (LOH) Algorithm in Genotyping Console 2.0
The Loss of Heterozygosity (LOH) Algorithm in Genotyping Console 2.0 Introduction Loss of erozygosity (LOH) represents the loss of allelic differences. The SNP markers on the SNP Array 6.0 can be used
More informationweek by week in the Training Calendar. Follow up graphics, sort, filter, merge, and split functions are also included.
Firstbeat ATHLETE The story Firstbeat ATHLETE is one of the most advanced software tools for analyzing heart rate based training. It makes professional level training analysis available to all dedicated
More informationDATE 2006 Session 5B: Timing and Noise Analysis
DATE 2006 Session 5B: Timing and Noise Analysis Bus Stuttering : An Encoding Technique To Reduce Inductive Noise In Off-Chip Data Transmission Authors: Brock J. LaMeres, Agilent Technologies Sunil P. Khatri,
More informationBangor University Laboratory Exercise 1, June 2008
Laboratory Exercise, June 2008 Classroom Exercise A forest land owner measures the outside bark diameters at.30 m above ground (called diameter at breast height or dbh) and total tree height from ground
More informationMonitor Instructions for Models: CHB-R6 CHB-UV6
Monitor Instructions for Models: CHB-R6 CHB-UV6 Power supply: output DC 9v/ 1000mA transformer adaptor TABLE OF CONTENTS Function Buttons ----------------------------------------------------------------------------
More informationResponse Time-Optimized Distributed Cloud Resource Allocation
Response Time-Optimized Distributed Cloud Resource Allocation Matthias Keller Holger Karl Computer Networks Group Universität Paderborn Minimizing response times Latency-critical service Interactive, emergency
More informationStatement of research interest
Statement of research interest Milos Hauskrecht My primary field of research interest is Artificial Intelligence (AI). Within AI, I am interested in problems related to probabilistic modeling, machine
More informationSONIMAGE HS1. Premium Portable Ultrasound
SONIMAGE HS1 Premium Portable Ultrasound EXIT SET F2 F1 User User Gain FREEZE PREMIUM PORTABLE ULTRASOUND DESIGNED FOR YOU Sonimage HS1 2 Sonimage HS1 is Konica Minolta s Premium Portable Ultrasound designed
More informationCOMPUTER INSTRUCTIONS
COMPUTER INSTRUCTIONS COMPUTER INSTRUCTIONS MODEL NUMBER: 9.35 2 A B M N C C D E F G H I J K L Calories/Calories Per Hour A B Time/Distance Preset Programs C D Motion Control Incline UP/DOWN E F Enter
More informationUsing the CFS Infrastructure
CHAPTER 13 The Cisco MDS SAN-OS software uses the Cisco Fabric Services (CFS) infrastructure to enable efficient database distribution and to foster device flexibility. It simplifies SAN provisioning by
More informationBe Strong. Model T9.16. Nautilus Commercial Series Treadmill Owner s Manual C H A N G I N G T H E G A M E I N H E A LT H A N D F I T N E S S NAUTILUS
C H A N G I N G T H E G A M E I N H E A LT H A N D F I T N E S S Model T9.16 Model T9.16 Be Strong. Nautilus Commercial Series Treadmill Owner s Manual C HA NG I N G T H E G A M E I N H EA LT H A N D F
More information10 Years Experience in Industrial Phased Array Testing of Rolled Bars
18th World Conference on Nondestructive Testing, 16-20 April 2012, Durban, South Africa 10 Years Experience in Industrial Phased Array Testing of Rolled Bars Josef MAIER 1 and Gerhard Ferstl 1, 1 Böhler
More information510Ic Owner s Manual CONSOLE OVERVIEW & GENERAL NOTES CONSOLE KEYPAD FUNTION
CONSOLE OVERVIEW & GENERAL NOTES A primary feature of the 510Ic is the ability to control and monitor the user s workout program via the brilliant LCD console. At any time, the user may view their speed,
More informationenvironment because of dynamically changing topology and lack of accurate and timely information,
ADAPTIVE FAULT LOCALIZATION IN MOBILE AD HOC BATTLEFIELD NETWORKS * Maitreya Natu, Adarshpal S. Sethi University of Delaware, Newark, DE, 19711 {natu, sethi}@cis.udel.edu ABSTRACT Future battlefield networks
More informationTDD: A New Engine for Mobile Broadband. GTI October, 2015
TDD: A New Engine for Mobile Broadband GTI October, 2015 GTI : Building Global TD-LTE Ecosystem Construct a robust ecosystem of TD-LTE Global TD-LTE Initiative is an operators-driven industry consortium
More informationWireless MIMO Switching with MMSE Relaying
Wireless MIMO Switching with MMSE Relaying Fanggang Wang, Soung Chang Liew, Dongning Guo Institute of Network Coding, The Chinese Univ. of Hong Kong, Hong Kong, China State Key Lab of Rail Traffic Control
More informationPRASHANTI JEYAMOHAN( インド ) 博士 ( 生命科学 ) 甲第 351 号 ( 甲生第 30 号 ) 本学学位規則第 3 条第 1 項該当
11 氏名 ( 本籍地 ) 学位の種類報告 学位記番号学位記授与の日付学位記授与の要件学位論文題目論文審査委員 PRASHANTI JEYAMOHAN( インド ) 博士 ( 生命科学 ) 甲第 351 号 ( 甲生第 30 号 ) 平成 25 年 9 月 25 日 本学学位規則第 3 条第 1 項該当 CARBON NANOTUBES AS A THERAGNOSTIC TOOL AGAINST
More information平成 23 年度特別研究報告 感性情報に着目したデザイン支援 ( 平成 24 年 10 月 16 日受理 ) Report for Special Research 2011 An Approach to Design by Kansei Information
広島国際学院大学研究報告, 第 45 巻 (2012),33~37 33 平成 23 年度特別研究報告 感性情報に着目したデザイン支援 趙 領逸 ( 平成 24 年 10 月 16 日受理 ) Report for Special Research 2011 An Approach to Design by Kansei Information Youngil CHO (Received October
More informationInformation document to CEPT PT1 on the possible impacts of SDL adoption in the MHz band
Jan. 15th, 2015 Information document to CEPT PT1 on the possible impacts of SDL adoption in the 2570-2620 MHz band To: CEPT ECC PT1 (Chairman Didier Chauveau) CC: Chairman, ETSI MSG TFES (Email: esa.barck@nsn.com);
More informationAdvanced Bayesian Models for the Social Sciences
Advanced Bayesian Models for the Social Sciences Jeff Harden Department of Political Science, University of Colorado Boulder jeffrey.harden@colorado.edu Daniel Stegmueller Department of Government, University
More informationExperiment HE-12: Targeted Exercise with Wireless Electrocardiogram (ECG)
Experiment HE-12: Targeted Exercise with Wireless Electrocardiogram (ECG) Get Your Cardio On Preparations for this experiment need to be reviewed with the students in advance. There will be exercise routines
More informationAn Orthodontics-Inspired Desynchronization Algorithm for Wireless Sensor Networks
An Orthodontics-Inspired Desynchronization Algorithm for Wireless Sensor Networks Pongpakdi Taechalertpaisarn Department of Computer Engineering Chulalongkorn University Bangkok, Thailand Email: pongpakdi.t@student.chula.ac.th
More informationTDD ratio estimation scheme for WiBro system with flexible TDD frame structure
TDD ratio estimation scheme for WiBro system with flexible TDD frame structure Eu-Suk Shim, Young-Il Kim, and Won Lyu Electronics and Telecommunications Research Institute, 218 Gajeongno, Yuseong-gu, Daejeon,
More informationCaring for your image
Caring for your image Based in Angouleme in Southwest France, ECM concentrates at this location the various talents of our highly skilled ultrasound know-how (Research and Development, Production, Marketing
More informationCSE 118/218 Final Presentation. Team 2 Dreams and Aspirations
CSE 118/218 Final Presentation Team 2 Dreams and Aspirations Smart Hearing Hearing Impairment A major public health issue that is the third most common physical condition after arthritis and heart disease
More informationUMBC REU Site: Computational Simulations of Pancreatic Beta Cells
UMBC REU Site: Computational Simulations of Pancreatic Beta Cells Sidafa Conde 1, Teresa Lebair 2, Christopher Raastad 3, Virginia Smith 4 Kyle Stern, 5 David Trott 5 Dr. Matthias Gobbert 5, Dr. Bradford
More informationTemporal Analysis of Online Social Graph by Home Location
Temporal Analysis of Online Social Graph by Home Location Shiori Hironaka Mitsuo Yoshida Kyoji Umemura Toyohashi University of Technology Aichi, Japan s143369@edu.tut.ac.jp, yoshida@cs.tut.ac.jp, umemura@tut.jp
More informationEgress Simulation Models and Their Application
Egress Simulation Models and Their Application S.M.V. Gwynne September, 2007 Rome Outline The importance of understanding human egress performance. Examples of incidents and associated investigations,
More informationOCCLUSION REDUCTION SYSTEM FOR HEARING AIDS WITH AN IMPROVED TRANSDUCER AND AN ASSOCIATED ALGORITHM
OCCLUSION REDUCTION SYSTEM FOR HEARING AIDS WITH AN IMPROVED TRANSDUCER AND AN ASSOCIATED ALGORITHM Masahiro Sunohara, Masatoshi Osawa, Takumi Hashiura and Makoto Tateno RION CO., LTD. 3-2-41, Higashimotomachi,
More informationMMCS Turkey Flu Pandemic Project
MMCS Turkey Flu Pandemic Project This is a group project with 2 people per group. You can chose your own partner subject to the constraint that you must not work with the same person as in the banking
More informationLab 5a Exploring Correlation
Lab 5a Exploring Correlation The correlation coefficient measures how tightly the points on a scatterplot cluster around a line. In this lab we will examine scatterplots and correlation coefficients for
More informationDEMAND AND CAPACITY MODELLING
DEMAND AND CAPACITY MODELLING How we used demand and capacity modelling to develop a robust and credible recovery plan Piers Young Deputy Chief Operating Officer (Elective Care) CONTENTS Brief history
More informationDerivation of Ratios for Female Urinals
for Female Urinals RM Harris 1 1. Contents 1. Contents... 2 2. Revision History... 2 3. Purpose... 3 4. Disclaimer... 3 5. Definitions... 3 6. Related Documents... 3 7. Summary... 3 8. Methodolgy... 3
More informationAssignment Question Paper I. 1. What is the external and internal commands and write 5 commands for each with result?
Subject:- INFORMATION TECHNOLOGY Max Marks -30 kvksa esa gy djuk vfuok;z gsa 1. What is the external and internal commands and write 5 commands for each with result? 2. Writes the features of UNIX Operating
More informationModel T9.14. Be Strong. NAUTILUS. Nautilus Commercial Series Treadmill Owner s Manual. Model T9.14
C H A N G I N G T H E G A M E I N H EA LT H A N D F I T N E S S Model T9.14 Be Strong. Model T9.14 Nautilus Commercial Series Treadmill Owner s Manual C HA NG I N G T H E G A M E I N H EA LT H A N D F
More informationMichael Hallquist, Thomas M. Olino, Paul A. Pilkonis University of Pittsburgh
Comparing the evidence for categorical versus dimensional representations of psychiatric disorders in the presence of noisy observations: a Monte Carlo study of the Bayesian Information Criterion and Akaike
More information(1) Experimental Verification of the Relationship between Pedal-to- Pedal Distance and Pedal Misapplication
(1) Experimental Verification of the Relationship between Pedal-to-Pedal Distance and Pedal Misapplication (2) Detection of Distress Caused by Pedal Misapplication and Its Application to Accident Prevention
More informationULTRASONIC ARRAY APPROACH FOR THE EVALUATION OF ELECTROFUSION JOINTS OF POLYETHYLENE GAS PIPING
ULTRASONIC ARRAY APPROACH FOR THE EVALUATION OF ELECTROFUSION JOINTS OF POLYETHYLENE GAS PIPING H. J. Shin 1, Y. H. Jang 1, J. R. Kwan 2, H. D. Lee 3 1 INDE System Co., Ltd., Suwon, Kyunggi-do, 440-746,
More informationHEAR BOTH SIDES OF THE STORY - WITH EVEN GREATER FLEXIBILITY
UN-AIDABLE HEAR BOTH SIDES OF THE STORY - WITH EVEN GREATER FLEXIBILITY USES THREE TIMES LESS BATTERY POWER THAN OTHER WIRELESS CROS SOLUTIONS SOUND IS TRANSMITTED FROM THE USER S UN-AIDABLE EAR TO THEIR
More informationReal-time computational attention model for dynamic scenes analysis
Computer Science Image and Interaction Laboratory Real-time computational attention model for dynamic scenes analysis Matthieu Perreira Da Silva Vincent Courboulay 19/04/2012 Photonics Europe 2012 Symposium,
More informationOverview. Video calling today. Who we are 8/21/12. Video Quality and Interoperability across Videophones Today and in the Future
Overview Video Quality and Interoperability across Videophones Today and in the Future Christian Vogler and Norman Williams Technology Access Program, Gallaudet University Who we are The state of video
More informationNANO: Network Access Neutrality Observatory
NANO: Network Access Neutrality Observatory Mukarram Bin Tariq, Murtaza Motiwala, Nick Feamster, Mostafa Ammar Georgia Tech 1 Net Neutrality 2 Net Neutrality 2 Net Neutrality 2 Example: BitTorrent Blocking
More information