Course Hero

Register now to access 7 million high quality study materials (What's Course Hero?) Course Hero is the premier provider of high quality online educational resources. With millions of study documents, online tutors, digital flashcards and free courseware, Course Hero is helping students learn more efficiently and effectively. Whether you're interested in exploring new subjects or mastering key topics for your next exam, Course Hero has the tools you need to achieve your goals.

75 Pages

Lecture21a

Course: CSE 498, Fall 2009
School: Oakland University
Rating:
 
 
 
 
 

Word Count: 3031

Document Preview

Challenges Three to Reliable Data Transport over Heterogeneous Wireless Networks Hari Balakrishnan Daedalus Group Department of Electrical Engineering and Computer Science University of California at Berkeley http://daedalus.cs.berkeley.edu http://www.cs.berkeley.edu/~hari Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Protocol Design for the Internet Internet invariants t t Heterogeneity Large...

Register Now
Search

Unformatted Document Excerpt

Coursehero >> Michigan >> Oakland University >> CSE 498

Course Hero has millions of student submitted documents similar to the one
below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.

Course Hero has millions of student submitted documents similar to the one below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.
Challenges Three to Reliable Data Transport over Heterogeneous Wireless Networks Hari Balakrishnan Daedalus Group Department of Electrical Engineering and Computer Science University of California at Berkeley http://daedalus.cs.berkeley.edu http://www.cs.berkeley.edu/~hari Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Protocol Design for the Internet Internet invariants t t Heterogeneity Large scale Protocols Applications Adaptation is crucial t t Importance of incremental deployment To design and implement adaptive protocols and applications for the Internet Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Motivation 25 Rapid growth 20 Cellular phones # of units/hosts (millions) 10 5 15 Sources: Ericsson, Inc. Matthew Gray, MIT Internet hosts 0 1993 1994 1995 1996 1997 But wireless data is floundering... t t Year Enormous heterogeneity Poor performance Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Goal: To make wireless devices first-class Internet citizens Wireless Heterogeneity Metricom Ricochet Lucent WaveLAN Regional-Area Metro-Area Cellular Digital IBM Infrared Packet Data (CDPD) Campus-Area Packet Radio In-Building Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Wireless Performance TechnologyRatedBandwidthTypical TCPThroughputIBMInfrar Goal: To bridge the gap between perceived and rated performance Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Methodology Deploy real networks Measure performance Identify and analyze problems Implement best solutions on real networks Evaluate performance Simulation (UCB/LBNL/VINT ns network simulator with several enhancements) Design solutions Evaluate solutions in simulation Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Structure TCP Background Challenge #1: wireless bit-errors t Solution: Berkeley Snoop Protocol Challenge #2: asymmetry and latency variation t Solution: TCP mods + link-level schemes Challenge #3: low channel bandwidths t Solution: Enhanced TCP loss recovery Conclusions Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Cellular Network Topology Base Station (BS) Internet Fixed Host (FH) Mobile Host (MH) Wireless Cell Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Internet Service Model Internet Router A best-effort network: losses & reordering can occur Need reliable data transport protocols t Web, file transfer, remote terminal, e-mail,... Efficient loss recovery Robust congestion and flow control Functions t t Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Transmission Control Protocol (TCP) 7 6 5 4 2 4 3 3 2 1 Cumulative Acknowledgments (ACK) Internet standard for reliable transport t 95% of all bytes, 90% of all packets (Thompson, et al.) New algorithms within this protocol framework Incremental deployment of modifications Flexible protocol framework t t Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT TCP Overview 1. Loss recovery 10 9 8 7 6 5 4 3 0 lost 1 0 1 1 1 1 2 Timeouts based on mean round-trip time (RTT) and deviation Fast retransmissions based on duplicate ACKs Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT 2. Congestion control [Jacobson88] t Window-based algorithm to determine sustainable rate t Upon congestion, reduce window t "ACK clocking" sends data smoothly TCP Dynamics 4.19E+06 Sequence number (bytes) 4.18E+06 Data 4.17E+06 Fast retransmission RTT Window Duplicate ACKs 4.16E+06 4.15E+06 ACKs 4.14E+06 4.13E+06 33.52 33.54 33.56 33.58 33.6 33.62 33.64 33.66 33.68 33.7 33.72 Time (s) Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Wireless Transport: The Three Challenges Preponderance of wireless bit-errors t Corruption vs. congestion losses Bandwidth asymmetry Latency variability Small windows Asymmetric effects t t Low channel bandwidths t Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Challenge #1: Wireless Bit-Errors Internet Router Loss Congestion Burst losses lead to coarse-grained timeouts 3 2 11 22 0 Loss ==> Congestion Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Result: Low throughput [CI95] Performance Degradation 2.0E+06 Sequence number (bytes) 1.5E+06 Best possible TCP with no errors (1.30 Mbps) TCP Reno (280 Kbps) 1.0E+06 5.0E+05 0.0E+00 0 10 20 30 40 50 60 Time (s) Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT 2 MB wide-area TCP transfer over 2 Mbps Lucent WaveLAN Conventional Approaches End-to-end Base Station ARQ/FEC Link-layer protocols [LC83] Adverse interactions with transport layer t Timer interactions [DCY93] t Interactions with fast retransmissions t Large end-to-end round-trip time variation Split connections [YB94,BB95] t Wireless connection need not be TCP Hard state at base station t Complicates mobility t Vulnerable to failures Violates end-to-end semantics Wired connection Wireless connection Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Our Solution: Berkeley Snoop Protocol Shield TCP sender from wireless vagaries t t Eliminate adverse interactions between protocol layers Congestion control only when congestion occurs Preserve TCP/IP service model: end-to-end semantics Is connection splitting fundamentally important? Is link-layer messaging fundamentally important? The End-to-End Argument [SRC84] t t Eliminate non-TCP protocol messages t Fixed to mobile: transport-aware link protocol Mobile to fixed: link-aware transport protocol Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: FH to MH 6 4 3 2 5 1 Snoop agent Base Station 1 FH Sender Snoop agent: active interposition agent t t t Snoops on TCP segments and ACKs Detects losses by duplicate ACKs and timers Suppresses duplicate ACKs from FH sender Mobile Host Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Cross-layer protocol design: snoop agent state is soft Snoop Protocol: FH to MH 1 Snoop Agent Base Station FH Sender Mobile Host Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: FH to MH 5 4 3 2 1 Base Station FH Sender Mobile Host Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: FH to MH 6 4 3 2 5 1 Base Station 1 FH Sender Mobile Host Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT 6 5 Snoop Protocol: FH to MH 4 3 2 1 Base Station Sender 3 2 1 2 Mobile Host Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: FH to MH 6 5 4 3 2 1 Base Station 4 Sender 3 2 Duplicate ACK ack 0 Mobile Host 1 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: FH to MH 6 5 4 3 2 1 6 Base Station 5 1 Sender ack 0 Retransmit from cache at higher priority ack 0 ack 0 4 3 2 Mobile Host 1 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: FH to MH 6 5 4 3 2 1 Base Station Sender ack 0 5 Suppress Duplicate Acks ack 4 1 4 3 2 Mobile Host 1 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: FH to MH 6 5 Clean cache on new ACK Base Station 6 Sender ack 4 5 1 4 3 2 ack 5 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: FH to MH 6 Base Station Sender ack 4 ack 5 6 1 5 4 3 2 ack 6 Mobile Host Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: FH to MH 7 9 8 Base Station Sender ack 5 ack 6 6 1 5 4 3 2 Active soft state agent at base station Transport-aware reliable link protocol Preserves end-to-end semantics Mobile Host Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Handling Mobility 5 4 Home Agent 3 2 2 1 Base Station (Snoop agent) Sender Base Station (Snoop agent) 1 1 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Handling Mobility 6 5 Home Agent 4 3 3 2 1 Base Station (Snoop agent) Sender Base Station (Snoop agent) 2 1 1 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: MH to FH Base Station 3 2 Receiver Caching and retransmission will not work Sender t t 0 1 2 Losses occur before packet reaches BS Congestion losses should not be hidden Solution: Explicit Loss Notifications (ELN) t t In-band message to TCP sender General solution framework Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: MH to FH 0 1 Receiver Base Station Sender Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: MH to FH 0 3 2 Receiver Base Station Sender 1 2 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: MH to FH Add 1 to list of holes after checking for congestion 1 2 5 4 3 Receiver 0 Base Station Sender ack 0 1 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: MH to FH 1 6 5 4 ack 0 Base Station ack 0 ack 0 Receiver 3 2 0 Duplicate ACKs Sender 1 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: MH to FH ELN marking 1 6 ack 0 ack 0 Base Station ack 0 ack 0 Receiver 5 3 2 4 0 Sender ack 0 ELN information on duplicate ACKs 1 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: MH to FH Retransmit on dup ACK + ELN No congestion control now 1 1 ack 0 ack 0 Base Station ack 0 ack 0 Receiver 6 3 5 2 4 0 Sender ack 0 ELN information on duplicate ACKs 1 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Snoop Protocol: MH to FH Clean holes on new ACK Receiver ack 6 Base Station 1 6 3 5 2 4 0 Sender Link-aware transport decouples congestion control from loss recovery Technique generalizes nicely to wireless transit links Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT End-to-End Enhancements ELN to decouple congestion from loss recovery 4 Selective ACKs ack 0 [sack 2] ack 0 [sack 2,4] 2 0 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Selective ACKS (SACK) for burst losses [FF96,KM96,MMFR96,B96] Snoop protocol: no changes to fixed hosts on the Internet Snoop Performance Improvement 2.0E+06 Sequence number (bytes) 1.5E+06 Best possible TCP (1.30 Mbps) Snoop (1.11 Mbps) TCP Reno (280 Kbps) 1.0E+06 5.0E+05 0.0E+00 0 10 20 30 40 50 60 Time (s) Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT 2 MB wide-area TCP transfer over 2 Mbps Lucent WaveLAN Performance: FH to MH 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0 500 1000 1500 2000 2500 Snoop+SACK Snoop Typical error rates SPLIT-SACK TCP SACK SPLIT TCP Reno Throughput (Mbps) Snoop+SACK and Snoop perform best Connection splitting not essential TCP SACK performance disappointing 1/Bit-error Rate (1 error every x Kbits) 2 MB local-area TCP transfer over Mbps Lucent WaveLAN Slides from Prof. Hari Balakrishnan's2job talks. Berkeley MIT Empirical Error Models 1.2 1 Data collected from Reinas Env. Monitoring Network Santa Cruz, CA Error duration Error-free duration 0.8 CDF 0.6 0.4 0.2 0 0 2 4 6 8 10 Duration (ln ms) Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT # of downloads in 1000 s Real-World Web Performance 3000 2500 2000 1500 1000 500 0 1 conn. 1 conn. 2 conns. 2 conns. 186 203 975 Reno 3 conns. 3 conns. 102 177 1033 SACK Snoop 4 conns. P-HTTP P-HTTP 966 985 3000 4 conns. 206 76 1085 Snoop performance improvement: 3X-6X over Reno & SACK Empirical wireless error model from real traces of Reinas wireless network, UC Santa Cruz Empirical Web workload model real from traces [Mah97] Reno SACK Snoop 170 179 849 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Benefits of TCP-Awareness Congestion Window (bytes) 60000 50000 40000 30000 20000 10000 0 0 10 Snoop LL (no duplicate ack suppression) 20 30 Time (sec) 40 50 60 70 80 30-35% improvement for Snoop: LL congestion window is small (but no coarse timeouts occur) Connection bandwidth-delay product = 25 KB Suppressing duplicate acknowledgments and TCP-awareness leads to better utilization of link bandwidth and performance Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Split-Connection Congestion Window Congestion Window (bytes) 60000 50000 40000 30000 20000 10000 0 0 20 40 60 80 100 120 Wired connection Wireless connection Time (sec) Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Wired connection does not shrink congestion window but wireless connection times out often, causing sender to stall Snoop Protocol Status BSD/OS implementation t Integrated with Daedalus handoff software [SBK97] Version 1 released 1996; Version 2 in beta Daily production use at Berkeley and UC Santa Cruz Several hundred downloads t Ports to Linux, FreeBSD, NetBSD Papers: MOBICOM 95, SIGCOMM 96, Trans. on Networking (Dec. 97) Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Summary: Wireless Bit-Errors Problem: wireless corruption mistaken for congestion Solution: Berkeley Snoop Protocol General lessons t t Lightweight soft-state agent in network infrastructure Guided by the End-to-End Argument Fully conforms to the IP service model Cross-layer protocol design & optimizations Transport Network Link-aware transport (ELN) Transport-aware link (Snoop agent at BS) Link Physical Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Challenge #2: Asymmetric Effects Asymmetric access technologies t t ADSL, (wireless) cable modems, DBS, etc. Low-bandwidth ACK channel [LM97, KVR98] Metricom's Ricochet, CDPD, etc. Adverse interactions between data and ACK flow Packet radio networks t t Problem: Imperfect ACK feedback degrades TCP performance Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT The Character of Asymmetry Router Forward Server Router ACK Client The network and traffic characteristics in one direction significantly affect performance in the other Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Bandwidth: 10-1000 times more in the forward direction Latency: Variability due to MAC protocol interactions Packet loss: Higher loss- or error-rate in one direction Bandwidth Asymmetry Problems Data 11 Router Data 10 Data 9 Forward Bottleneck Router Data 8 Server ACK 0 Client 1 2 3 4 5 6 7 1 4 3 2 6 7 5 1. Acks arrive slowly (large buffer) 2. Acks are dropped (small buffer) 3. Acks are queued behind data packets Data Data 1 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Lakshman & Madhow 97 Kalampoukas et al. 97 Balakrishnan et al. 97 Ack flow Hybrid Wireless Cable Measurements TCP Throughput (Mbps) 6 5 4 3 2 1 0 0 20 40 60 10 Mbps Ethernet 28.8 C-SLIP 9.6 C-SLIP 9.6 SLIP 80 100 120 28.8 SLIP 140 160 180 200 Return channel speed and latency affects performance Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Socket Buffer Size (KB) Latency Asymmetry: Packet Radio Networks Fixed Host FH Ethernet Radios ER PT CTS RTS PT Mobile Host PT MH Modem PR Internet GW ER PT PT ER PT Poletop Radios PT Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Half-duplex radios Synchronization before communication Packet Radio Networks Fixed Host FH Ethernet Radios ER PT Ack Dat a PT Mobile Host No response PT RTS MH Modem PR Internet GW ER PT PT ER PT Poletop Radios Exponential PT backoff Problem: Large and variable communication latency Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Problem: Large Round-Trip Time Variations Example: Metricom Ricochet Wireless Network 300000 250000 200000 150000 100000 Sequence Number trace RTT Estimate 6000 5000 Fast retransmissions Timeouts 4000 3000 2000 1000 0 Sequence Number (bytes) 0 0 20 40 50000 RTT Estimate (msec) 1 3 5 Time (sec) 60 80 100 Sample number 7 9 11 13 15 17 19 Mean rtt = 2.45s, std deviation = 1.5s long timeout! Long idle periods after multiple losses (~ 20 Kbps) In contrast, UDP throughput = 50-64 Kbps ACK flow Hari Balakrishnan's Slides from Prof. affects data latency job talks. Berkeley MIT Solutions Problems arise because of imperfections in the ACK feedback Reduce frequency of acks t t ACK Filtering (AF) ACK Congestion Control (ACC) Sender Adaptation (SA) ACK Reconstruction (AR) Handle infrequent acks t t General solution approach for asymmetric situations Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT ACK Filtering (AF) Router Forward Router Server 1 3 5 7 9 11 13 Client Purge all redundant, cumulative ACKs from constrained reverse queue Used in conjunction with sender adaptation or ACK reconstruction Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT ACK Congestion Control (ACC) Data 20 Data 21 Data 19 Data 22 Router Forward 8 16 10 14 18 Client Server Delack factor = 2 Adaptive extension of TCP delayed ACKs based on Slides congestion Hari Balakrishnan's jobor sender from Prof. feedback from router talks. Berkeley MIT ACK Congestion Control (ACC) Data Data Data Data Router Forward Data Client 12 Server Delack factor = 2 RED [FJ93] marking of ECN bit [F94] (Explicit Congestion Notification) 22 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT ACK Congestion Control (ACC) Data Echo ECN marking to receiver Data 40 Data Router Forward Data Client 22 Server Delack factor = 2 Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT ACK Congestion Control (ACC) Data 42 Data 41 Data 43 Router Forward Data 40 36 40 Client Delack factor = 4 Server Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT Sender Adaptation (SA) Infrequent ACKs cause slow window growth Sender tends to be bursty Router Forward Server 1 9 15 Client 1. cwnd += 8 cwnd += 8/cwnd 2. Increment window by This reduces amount ofProf. ack'd Balakrishnan's job burstiness data Hari Slides from talks. Berkeley MIT ... Regulation: pace packets out at rate estimated by cwnd/srtt 19 20 21 22 ACK Reconstruction (AR) Forward Server 1 3 1 5 7 9 3 5 7 9 11 13 Client ACK filter ACK reconstructor Regenerates ACKs at other end of reverse channel Shields sender from large gaps in ack sequence AR rate determined by t t Slides from Prof. Hari Balakrishnan's job talks. Berkeley MIT input ACK rate target ACK spacing Bandwidth Asymmetry Performance t t TCP transfers in the forward direction alone Maximum window size 100 KB; no losses on forward path 10 8 6 4 2 Throughput (Mbps) ...

Find millions of documents on Course Hero - Study Guides, Lecture Notes, Reference Materials, Practice Exams and more. Course Hero has millions of course specific materials providing students with the best way to expand their education.

Below is a small sample set of documents:

Penn State - TRDEV - 418
Basic Communication Chapter 6CommunicationWas there communication as I read the Jabberwocky? How did the communication come about? Were you familiar with all the words? If not, how did you make meaning of them?What emotions did you feel? Why? What infl
Penn State - MKTG - 572
1995 7888 4320 000 000001 00023Copyright 2003 by The McGrawHill Companies, Inc. All rights reserved. C H A P T E R1995 7888 4320 000 000001 000239Descriptive Research Designs: Survey Methods and Errors1234 0001 897251 000009-2Copyright 2003 by The
Arkansas Little Rock - FIN - 3310
Chapter ThirteenReturn, Risk, and the Security Market Line 2003 The McGrawHill Companies, Inc. All rights reserved.13.2Key Concepts and Skills Know how to calculate expected returns Understand the impact of diversification Understand the systematic ri
Delaware - CMSC - 411
MemoryHierarchies: BringingItAllTogetherCMSC411/ComputerArchitectureTheseslidesandallassociatedmaterialare 2003byJ.Sixandareavailableonlyfor studentsenrolledinCMSC411.Fortunateishe,whocanunderstandthecausesofthings. CMSC411ComputerArchitecture/2003J.S
Cal Poly Pomona - HRT - 383
Teamwork Scoring Rubric HRT 383 Peer Evaluation SheetEvaluator's Name: _ , _ Section 01 02 Last First"We are not born knowing how to work as a team, and teamwork can be difficult to learn. Perhaps this is one of the reasons that employers value so highl
UCF - EEL - 4768
EEL 4768 Computer System Design 2Lecture 9: Memory Systems (cont.)Adapted from D.A.Patterson, UCBThe Art of Memory System DesignWorkload or Benchmark programsProcessorreference stream <op,addr>, <op,addr>,<op,addr>,<op,addr>, . . . op: i-fetch, read
University of Dayton - PSY - 217
Theory and Data in Psychology What is a theory? Theories at multiple levels How we know when we're wrong Hypotheses/Predictions -> Confirmation/Disconfirmation Can't "prove" a theory Falsifiability Examples of Theory-Testing Process1Introduction Fixin
Concordia Moorhead - PPT - 306
Chapter 10Muscle Tissue(Mostly Skeletal Muscle)Muscle tissue: skeletal bones cardiac heart smooth "hollow organs"Muscle tissue: skeletal muscle functions move skeleton maintain balance/posture support soft tissues guard entrances/exits maintain body t
UMBC - CS - 203
MoresophisticatedcountingAdvancedCounting Textsections 4.5&4.6Fall 2003 CMSC 203 Discrete Structures 1 Theorem: Suppose a set A has n elements and is partitioned by the collection cfw_A1, A2, ., Ap, where each partition set has m elements. Then: p = n/
CSU Sacramento - PHIL - 154
Stainton, Robert Philosophical Perspectives on Language Chapter 7 The Language Of ThoughtConcepts/Terminology/Skills Language of thought Mentalese Intentional states Propositional attitudes Cognitive science Analogue representation Formulae Propositional
Davidson - ECO - 105
Name:Statistics Economics 204, Fall 2002Davidson College Mark C. FoleyReview # 2Suggested SolutionsProblem 1 Let three random variables have the following distributions: X ~ N (4,25) , Y ~ N (4,64) , and W ~ N ( 2,25) . (a) Calculate the probability
CSU San Marcos - NFS - 362
Chapter 16 Life Cycle Nutrition: Infancy, Childhood and AdolescenceNutrition During Infancy I. Energy and Nutrient NeedsA. Energy Intake and Activityst1. Fastest rate of growth occurs in 1 yeara. Weight doubles in 4 mo., triples in 12 mo.2. BMR twic
CSU LA - ACCT - 211
Profitability AnalysisAppendix BMcGraw-Hill/IrwinCopyright 2008, The McGraw-Hill Companies, Inc.B-2Absolute ProfitabilityAbsolute profitability measures the impact on the organization's overall profits of adding or dropping a particular segment such
Wisc Green Bay - NUTSCI - 486
Chapter 40Medical Nutrition Therapy for Cancer Prevention, Treatment, and RecoveryCancersWhen cells divide and reproduce abnormally and have the potential to spread throughout the body, crowding out normal cells and tissues One third of the cancer dea
Cal Poly Pomona - EWS - 425
Technology VS ScienceBrought to You Exclusively by:Richard Lum, Stephanie Logan, Jerry Clyde, Diana Seif, & Edgar KhachatryanNECESSITY = THE MOTHER OF INVENTION Team ABCDE 1Technology Defined. Technology = Doing Science=UnderstandingTeam ABCDE2Sci
San Diego State - CSC - 105
Introduction to the Internet and The World Wide WebB. Tudin Information Technologies04/20/091What on Earth is the Internet?The Internet is the largest network of computers in the world using the TCP/IP protocols. (Network of Networks)B. Tudin Inform
TAMU Commerce - BA - 302
Chapter 3Descriptive Statistics: Numerical MethodsMcGraw-Hill/IrwinDescriptive Statistics3.1 Describing Central Tendency 3.2 Measures of Variation 3.3 Percentiles, Quartiles and Box-andWhiskers Displays 3.4 Covariance, Correlation, and the Least Squar
Ill. Chicago - EECS - 265
Web code: Last name: First name:ECE 267, Fall 2006 Semester HW2 due Wednesday, Sept. 13, 2006.Grading stamp1. What is (17.2)10 converted into radix 2? 2. What is (3.8)16 converted into radix 10? 3. What is ( 10 written in 8-bit binary signed integer co
National Taiwan University - COB - 521
Financial Accounting IWilliam F. BentzAccounting & MISW. F. BentzEarly Accounting for ASSETSqEarly accounting amounted to listing one's assets.W. F. BentzIncome on a Cash BasisEarly measures of income were cash based. Income is equal to cash rece
Kenyon - MATH - 102
Price vs. Length : How Much Are You Getting For Your Money?Goal: To compare runtime of a CD to its cost to see if there was a connection between how much music was on a disc and how much I was paying for it. I then used this data to see if the style of m
Southern New Orleans - CS - 4350
More on Thread APISpring/2002Distributed Software Engineering C:\unocourses\4350\slides\DefiningThreads1public void interrupt(); Receiving thread is interrupted. public boolean isInterrupted() Tests whether this thread has been interrupted. public sta
Idaho - SOILS - 700
1.E+01Relative Hydraulic Conductivity0.4 0.6 0.8 1.0 1.21.E-011.E-031.E-051.E-07 0.0 0.1 1.0 10.0 100.0 1000.0 10000.0Degree of Saturation- Chemical Potential [J/kg]TRIANGLE 60 1.05 0.43 2.0546 6.158 5.970E-04 8.000E+03 5.000E-09 4.578E-04GAMMA
U. Houston - ACCT - 5337
CHAPTER 9Standard Costing:A Functional-Based Control ApproachLEARNING OBJECTIVESAfter studying this chapter, you should be able to: 1. Describe how unit input standards are developed, and explain why standard costing systems are adopted. 2. Explain the
North Texas - CAS - 1710
Physics 1710-Warm-up QuizA jet ski expels water at a rate of 1440. liters per minute at a velocity of 45.00 m/s. What thrust does it produce? 1. 2. 3. 4. 45.00 N 1440. N 1080. N 14112. N0% 0 of 110 Answer Now !0%10%20%30%41 212 223 234 245
North Texas - CAS - 1710
Physics 1710-Warm-up Quiz0A jet ski expels water at a rate of 1440. liters per minute at a velocity of 45.00 m/s. What thrust does it produce? 1. 2. 3. 4. 45.00 N 1440. N 1080. N 14112. N42% 59 of 140Answer Now !086%8% 0%1 2 35%441 6142 6243
Iowa State - CPRE - 558
Scheduling tasks with precedence relationsConventional task set cfw_T1, T2 SchedulerT1T2task set with precedence constraintsModify task parameters in order to respect precedence constraintsSchedulerCprE 458/558: Real-Time Systems (G. Manimaran)1M
RIT - IST - 19971
Fundamentals of TelecommunicationsWeek 6-7: Internets & IntranetsElizabeth Lane Lawley, InstructorICSA 733Week 6-71What is the Internet? Collection of interconnected networks using TCP/IP protocols to communicate with each other No single owner or
Clarkson - ME - 411
INTRODUCTION TO HEAT TRANSFER (ME 411)Summer 2003Lecture Number IVAli R. MazaheriCourse webpage: www.clarkson.edu/~mazahear/MAIN/ME411Department of Mechanical and Aeronautical Engineering Clarkson UniversityTwo-Dimensional Heat Conduction2 T T + 2
Berkeley - EE - 219
Delay-Optimal Technology Mapping by DAG CoveringYuji Kukimoto Robert K Brayton Prashant SawkarPresented by Bret Victor, 4/5/00Abstract An algorithm for minimal-delay library-based technology mapping Subject graphs can be mapped directly as DAGs, witho
Iowa State - CS - 586
C hapte 4: Ne r twork Laye rC hapte goals: r unde rstand principle be s hind ne twork laye se s: r rvice r routing (path se ction) le r de aling with scale r how a route works r r advance topics: Multicast routing, I Pv6, m d obility instantiation and im
Yale - CS - 112
1CS112 Introduction to ProgrammingLectures 34-35: Data StructuresZhong Shao Department of Computer Science Yale University 314 AK Watson, Phone: 432-6828 Email: shao@cs.yale.edu2Introduction Dynamic data structures Grow and shrink at execution time
Oakland University - CSE - 498
Outline Putting it all together Upgrading student desktops to GigE Why? Why not?Jun 22, 20094/598N: Computer NetworksWeb Server Web server for ND: What Network Interface Card (NIC) would you use, 100 Mbps Ethernet, 1 Ge? Would having multiple NICs h
Washington University in St. Louis - CSE - 573
CSE 573S: Networking ProtocolsOverview Instructor: Manfred GeorgAcknowledgmentsMany slides are from Dina Katabi who stated:"Some slides are from lectures by Nick Mckeown, Ion Stoica, Frans Kaashoek, Hari Balakrishnan, and Sam Madden"Other slides are
University of Toronto - ECE - 1770
Distributed Hash TablesDavid Tam Patrick PangPresentation Outline What is DHT (Distributed Hash Table)? Why DHTs? Applications How lookup works? Alternatives to DHTs Performance Routing Performance Load Balancing Security Routing Attack Security Incons
MS Women - ED - 497
TeachersDiscoveringComputersIntegratingTechnologyand DigitalMediaintheClassroom 4thEditionChapter 8Security Issues, Ethics, and Emerging Technologies in EducationChapter Objectives Identify security risks that threaten home and school computers Descr
Arkansas Fort Smith - CSC - 350
Chapter 6 Multimedia NetworkingA note on the use of these ppt slides:Were making these slides freely available to all (faculty, students, readers). Theyre in powerpoint form so you can add, modify, and delete slides (including this one) and slide conten
Rose-Hulman - AY - 0506
Mechanical EngineeringME 510Gas DynamicsCatalog Description Introduction to the dynamics of a compressible flow. Equations of motion for subsonic and supersonic flow. Nozzle flow. Normal and oblique shock waves. Prandtl-Meyer flow. Steady and unsteady,
Washington - TC - 518
Product StatementThe Fujitsu UScan self-checkout system is in use at Fred Meyer stores. The system is an alternative to the traditional check-out counter. It consists of a kiosk touch screen, a UPC-code scanner/weighing station, a bagging station, and me
UCF - COT - 5937
Introduction to Cryptography Summer '03 Homework #1 Solutions1) The hint pointed to the fact that there were no e's in the passage, since odd and atypical don't have the letter e, but peculiar and strange both do.) Trying to plug in e for the most common
Penn State - RAL - 242
1Abortion Do you think abortion should be allowed or not? cases of abortion are viewed to be legal. SomeIt is thoughtthat abortion should be allowed in certain circumstances such as rape and if severe medical complications occur during the mother's preg
Cornell - ECON - 102
CHAPTER23Economic Growth in Developing and Transitional EconomiesPrepared by: Fernando Quijano and Yvonn Quijano 2004 Prentice Hall Business PublishingPrinciples of Economics, 7/eKarl Case, Ray FairEconomic Growth in Developing and Transitional Eco
CSU Northridge - CCS - 53520
A.P. Chapter 13 Outline: Chemical Kinetics Chemical Kinetics: the study of the speeds of reactions and the nanoscale pathways or rearrangements by which atoms and molecules are transformed from reactants to products. I. Reaction Rate A. Homogeneous reacti
N.E. Illinois - ECN - 2802
Chapter 6Consumer Choice and Demand 2006 Thomson/South-Western1Utility AnalysisUtility: the sense of pleasure, or satisfaction, that comes from consumption Tastes: preferences for different goods and services likes and dislikes2Total and Marginal U
Fayetteville State University - ECO - 2023
Taking the Nations PulseFull Length Text Part: 3 Macro Only Text Part: 3 Chapter: 7 Chapter: 7To Accompany Economics: Private and Public Choice 10th ed. James Gwartney, Richard Stroup, Russell Sobel, & David Macpherson Slides authored and animated by: J
Kent State - BUSINESS - 24163
Decision MakingCHAPTER SIXCopyright 2005 by South-Western, a division of Thomson Learning. All rights reserved1 CHAPTER SIXWhat Would You Do?You are the new CEO of Baxter Healthcare Corporation H H You want to expand the hemodialysis business But, wo
Dallas - LCK - 016000
WRAPUP UK: Expansion of judicial review three key RIGHTS areas (local to executive) Politicians (nonjudicial actors) using courts judicial forms/methods/ processes Judicialization occurs many contexts ARTICLES: Creation of new court in 1971 (to deal with
Purdue - AAE - 251
FTP information for AAE251 These instructions are for FTP access to your team directory in the `aae251su' account from Grissom (ecn) computers. You should all have access to your ecn account. Within the aae251su account there is a directory called `groups
Penn State - KMG - 5189
Redesign of a Single Use CameraEDSGN 100 Team 8 Kara Gallo, Catie Martel, Nick Blymiller, Jason Clark, Ashley Pachter October 2, 2007Overview Camera Packaging and Parts Analysis Battery Data and Reuse Survey Current Recycling Practices New Design
North Texas - CAS - 1710
Physics 1710-Warm-up Quiz0Why does a diver rotate faster when she tucks in her arms and legs?Answer Now ! A. B. C. D. E. She increases her angular momentum. She increases her moment of inertia. She decreases her moment of inertia. She pushes against he
Binghamton - CS - 527
Mobile Ad Hoc Routing (IV) Uses material from tutorial by Nitin Vaidya1Last Timeo oooo oFinished Cache optimization for reactive protocols Preemptive Routing o Discussion: optimizing reactive protocols Link Reversal algorithms and TORA o Try to loc
Cal Poly Pomona - EVALUATION - 550
4. Analysis of sections INFLUENCES ON NEIGHBORHOOD (A5-A17; G60) a. Questions Influence of litter and trash on neighborhood (Becky) Influence of public drunkenness (Becky) Influence of young people hanging out and making noise (Becky) Influence of young p
Penn State - DMS - 5020
1 David Salberg There are many things that today's society worries about; television, role models, etc. The biggest ones are those that affect the children. One that many people overlook is video games, a national past time in almost every home. It is cle
Ball State - ITDPT - 203
Mechanical FasteningHold the parts without the use of outside assembly devices. Internal mechanical forces of the material position and hold parts in place.Physical Structures Parts can be designed to interlock. These physical structures include: 1. S
Siena - CSIS - 116
Chapter 9Electronic Commerce and Electronic Business1What is Electronic Commerce?Definition Electronic Commerce (e-commerce): Use of communication networks, including the public Internet, to conduct commercial transactions between businesses or with
UCF - COT - 3100
Some Sample Probability Questions1) A box of screws contains 5% defective screws. How many screws have to be chosen at random before there's a greater than 50% chance that at least one is defective? (Assume that after pulling several screws that the prob
Yale - RP - 269
STATE YEAR DISTRICT ARICE QRICE YRICE AMAIZE QMAIZE YMAIZE ABAJRA QBAJRA HARYANA1996 HISAR 55 160 2909 . . . 84 80 HARYANA1996 SIRSA 26 87 3346 . . . 5 5 HARYANA1996 BHIWANI . . . . . . 169 114 HARYANA1996 ROHTAK 17 38 2235 . . . 51 20 HARYANA1996 SONEPAT
Berkeley - BUSINESS - 280
BA 180/BA 280 Midterm Fall 2000 Suggestive Solutions for Question 1(a) Since Cap Rates are market driven, it is possible that each city for "identical physical office space" will, over the long run, have different expectations for cash flows (rental) gro
Rutgers - MS - 501
Sverdrup's theory of the oceanic circulation (Started 10/30 finish 11/2)We can arrive at a very robust version of the key features of the Sverdrup balance without needing to make the 1 layer assumption, or consider the details of the vertical stratificat
Purdue - AAE - 251
Solving the Equilibrium ODE in the Troposphere: (giving a lot of detail, as in MA266): 1) List all equations a) dP= - gdh : Equilibrium b) T = Ts - h : Temperature Profile (empirical) c) P = RT : Perfect Gas Law2) Known Constants: g, Ts, -dT / dh , R 3)
Auburn - ENG - 3600
CHEN3600 Problem Solving Fall 2007 Step 4 Systematic Approach to Problem SolvingIn order to solve a problem systematically, one needs to be constantly aware of the state of the solution, that is, whether or not sufficient information and relationships ha
University of Wisconsin - Fond du Lac - PED - 209
Chapter 8 Vitamins Vitamins: vital dietary components Definition- Essential organic substances needed in small amounts in the diet for normal function, growth and maintenance of the body Yield no energy to the body, they often participate in energy-yield