Hard Disk Drive Traffic Model for TGad

Slides:



Advertisements
Similar presentations
Submission doc.: IEEE /1214r1 September 2014 Leif Wilhelmsson, Ericsson ABSlide 1 Impact of correlated shadowing in ax system evaluations.
Advertisements

Doc.: IEEE /0604r1 Submission May 2014 Slide 1 Modeling and Evaluating Variable Bit rate Video Steaming for ax Date: Authors:
Doc.: IEEE /0709r2 Submission July 2009 Carlos Cordeiro, IntelSlide 1 Lightly Compressed Video Traffic Modeling Date: Authors: NameAffiliationsAddressPhone .
Lin Yingpei (Huawei Technologies) doc.: IEEE /1438r0 Submission November 2013 Slide 1 Traffic Observation and Study on Virtual Desktop Infrastructure.
Submission doc.: IEEE 11-14/0070r0 Jan 2014 Josiam et.al., SamsungSlide 1 Joint MAC/PHY Evaluation Methodology Date: Authors:
Doc.: IEEE /0364r1 SubmissionEldad Perahia, Intel CorporationSlide 1 Date: Authors: Antenna Array Gain from Measured Data for n/ac.
Doc.: IEEE /0090r0 SubmissionMartin Jacob, TU Braunschweig January 2010 Slide 1 Modeling the Dynamical Human Blockage for 60 GHz WLAN Channel.
27th, Nov 2001 GLOBECOM /16 Analysis of Dynamic Behaviors of Many TCP Connections Sharing Tail-Drop / RED Routers Go Hasegawa Osaka University, Japan.
Doc.: IEEE /1222r1 Submission November 2009 Eldad Perahia, Intel CorporationSlide 1 Hard Disk Drive Traffic Model for TGad Date: Authors:
Content caching and scheduling in wireless networks with elastic and inelastic traffic Group-VI 09CS CS CS30020 Performance Modelling in Computer.
Submission doc.: IEEE 11-14/0866r0 July 2014 Johan Söder, Ericsson ABSlide 1 Traffic modeling and system capacity performance measure Date:
1 An Analytical Model for the Dimensioning of a GPRS/EDGE Network with a Capacity Constraint on a Group of Cells r , r , r Nogueira,
Submission doc.: IEEE /1214r0 September 2014 Leif Wilhelmsson, Ericsson ABSlide 1 Impact of correlated shadowing in ax system evaluations.
Doc.: IEEE /2446r0 Submission September 2007 Eldad Perahia, Intel CorporationSlide 1 MIB Attributes for 40 MHz Scanning in 2.4 GHz Date:
Doc.: IEEE /0535r0 Submission May 2008 Thomas Kenney, Minyoung Park, Eldad Perahia, Intel Corp. Slide 1 PHY and MAC Throughput Analysis with 80.
Doc.: IEEE /0499r1 Submission May 2009 Eldad Perahia, Intel CorporationSlide 1 Simulation Scenario Floor Plans Date: Authors:
Accelerating Peer-to-Peer Networks for Video Streaming
Impact of LTE in Unlicensed Spectrum on Wi-Fi
TGad interference modeling for MAC simulations
TGad Process Overview Date: Authors: January 2009
IEEE aj Process Overview
Round Robin vs. Fair Queuing
April 2007 doc.: IEEE /0570r0 July 2010 Response to ITU-R Liaison to IEEE on “Multiple Gigabit Wireless Systems in frequencies around 60 GHz”
Proposal for TGad Evaluation Methodology
TGad March 2010 Closing Report
TGad September 2010 Closing Report
AP Service Load: Improved Definition
Lecture on Markov Chain
Performance Questions
April 2007 doc.: IEEE /0570r0 July 2010 Response to ITU-R Liaison to IEEE on “Multiple Gigabit Wireless Systems in frequencies around 60 GHz”
Revisions of Home Network Simulation Scenario with OBSS
Revisions of Home Network Simulation Scenario with OBSS
Revisions of Home Network Simulation Scenario with OBSS
Home Network Simulation Scenario with OBSS
TGad Channel Model Update
New Insights from A Fixed Point Analysis of Single Cell IEEE 802
TGad July 2009 Closing Report
Record and Playback PHY Abstraction for n MAC Simulations
Hawaii Liaison Report Date: Authors: September 2007
EE 122: Network Performance, Queueing Theory, Evaluation
OFDMA performance in 11ax
TGad September 2009 Closing Report
TGad November 2009 Closing Report
TGad January 2010 Closing Report
Taipei Liaison Report Date: Authors: January 2008
Chapter 11 I/O Management and Disk Scheduling
IEEE aj Process Overview
Update on “Channel Models for 60 GHz WLAN Systems” Document
IEEE aj Process Overview
TGad May 2009 Closing Report
A Fair Scheduling Scheme for HCF
NIST model output Date: Authors: May 2010 Month Year
VHT SG November 2008 Closing Report
CSE 550 Computer Network Design
Record and Playback PHY Abstraction for n MAC Simulations
TGad March 2011 Closing Report
Doppler Measurements for Mobile Devices
Investigation of upsampling techniques for TGac Channel Model
More Simulations on Secondary CCA
TGad May 2011 Closing Report
Response to Coexistence Presentations
Month Year doc.: IEEE /0578r0 May 2016
Proposed Addition to Evaluation Methodology
40 MHz Vs 20 MHz for video Date: Authors: July 2009
TGad January 2009 Closing Report
Simulation Scenario Floor Plans
TGad Task Group Document Open Items
Proposal for TGad Evaluation Methodology
Modeling and Evaluating Variable Bit rate Video Steaming for ax
AP Shut Out Neighborhood Effect
Presentation transcript:

Hard Disk Drive Traffic Model for TGad April 2007 doc.: IEEE 802.11-07/0570r0 November 2009 Hard Disk Drive Traffic Model for TGad Date: 2009-11-17 Authors: Eldad Perahia, Intel Corporation Eldad Perahia, Intel Corporation

November 2009 Background TGad Evaluation Methodology document includes a “hard disk file transfer” application Enterprise cubicle scenario File transfer wirelessly between laptop and hard drive Traffic model is TBD Eldad Perahia, Intel Corporation

Measured Data and Modeling November 2009 Measured Data and Modeling Hard Disk test in the PCMark Vantage benchmarking suite used to capture sample trace Specific subtest captured was “Importing pictures to Windows Photo Gallery” Reading a large collection of picture files from the hard drive Data captured is the transactions size and request time of the transactions Transactions are all reads from the hard drive Requests can be overlapping in time, i.e. a subsequent request arrives before the previous transaction is complete Model of the offered load of the transaction (packet size) and inter-arrival time of the requests is created with discrete random distributions Eldad Perahia, Intel Corporation

November 2009 Transaction model A transaction consists of a READ request from host to drive for a specific block of data Followed by the data transfer from drive to host For the purpose of the model: READ request is a short (256B) packet sent from host to drive Data transfer is instantaneous delivery of 4kB – 128kB to transmit queue in drive to host direction Assume fixed 1ms delay between receipt of READ request and data offered In practice, delay varies based on seek latencies of drive Not a significant component of the model READ (256B) Data (4kB-128kB) Inter-arrival time Inter-arrival time Eldad Perahia, Intel Corporation

Measured Transaction Size November 2009 Measured Transaction Size Most of the transaction are 64 Kbytes and 128 Kbytes Eldad Perahia, Intel Corporation

Measured Inter-Arrival Time of Requests November 2009 Measured Inter-Arrival Time of Requests Large grouping of inter-arrival times between 40usec and 1.1msec Inter-arrival times of requests of 40usec indicate overlapping requests Also high frequency of inter-arrival times of 50msec Probably due to operating system processing delay Eldad Perahia, Intel Corporation

Data Rate of Measured Data based on Transaction Completion Time November 2009 Data Rate of Measured Data based on Transaction Completion Time Maximum data rate equals 2.2 Gbps Average data rate (total transaction/total time) = 31 Mbps Eldad Perahia, Intel Corporation

November 2009 Model Step 1: Compute sequence of inter-arrival times of transaction requests Step 2: Compute corresponding sequence of transaction data sizes Eldad Perahia, Intel Corporation

Transaction Data Size Model November 2009 Transaction Data Size Model Discrete Random Variable distribution Eldad Perahia, Intel Corporation

Inter-arrival Times of Transaction Requests Model November 2009 Inter-arrival Times of Transaction Requests Model Discrete Random Variable distribution Eldad Perahia, Intel Corporation

Simulated Inter-Arrival Time of Requests November 2009 Simulated Inter-Arrival Time of Requests Simulated “time” is derived by a cumulative sum of the inter-arrival time Eldad Perahia, Intel Corporation

Simulated Transaction Size November 2009 Simulated Transaction Size Simulated average data rate over the entire period is equal to 30Mbps Eldad Perahia, Intel Corporation

November 2009 Conclusion Incorporate described hard disk drive model into Evaluation Methodology document Eldad Perahia, Intel Corporation