Multi-band OFDM Physical Layer Proposal Response - PowerPoint PPT Presentation

Loading...

PPT – Multi-band OFDM Physical Layer Proposal Response PowerPoint presentation | free to download - id: 6c045e-YzdlN



Loading


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation
Title:

Multi-band OFDM Physical Layer Proposal Response

Description:

Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs) – PowerPoint PPT presentation

Number of Views:72
Avg rating:3.0/5.0
Slides: 170
Provided by: ChuckBr8
Learn more at: http://grouper.ieee.org
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: Multi-band OFDM Physical Layer Proposal Response


1
Project IEEE P802.15 Working Group for Wireless
Personal Area Networks (WPANs) Submission Title
Multi-band OFDM Physical Layer Proposal Response
to no Voters Date Submitted 14 September
2003 Source Presenter see page 2 for the
complete list of participating authors and their
companies Re This submission is in response
to the IEEE P802.15 Alternate PHY Call for
Proposal (doc. 02/372r8) that was issued on
January 17, 2003. Abstract This document gives
modifications and more details for Multi-band
OFDM proposal for IEEE 802.15 TG3a
(doc.03/267/r2). Purpose To address the
concerns raised by the no voters in the July03
meeting Notice This document has been prepared
to assist the IEEE P802.15. It is offered as a
basis for discussion and is not binding on the
contributing individual(s) or organization(s).
The material in this document is subject to
change in form and content after further study.
The contributor(s) reserve(s) the right to add,
amend or withdraw material contained
herein. Release The contributor acknowledges and
accepts that this contribution becomes the
property of IEEE and may be made publicly
available by P802.15.
2
This contribution is authored by
  • Alereon - Vern Brethour, Martin Gravenstein, Joy
    Kelly, Tom Matheney, Kevin Shelby
  • General Atomics, Photonics Division - Naiel
    Askar, Susan Lin
  • Intel Corporation - Chuck Brabenac, Jeff
    Foerster, Dave Leeper, Srinivasa Somayazulu,
    Stephen Wood
  • Mitsubishi Electric - Andy Molisch, Yves-Paul
    Nakache, Jin Zhang
  • Philips - Charles Razzell
  • Samsung Electronics - Seung Young Park, Yongsuk
    Kim
  • Staccato Communications - Roberto Aiello, Nishant
    Kumar, Lars Mucke, Torbjorn Larsson, Larry Taylor
  • ST Microelectronics - Ljubica Blazevic, Glyn
    Roberts
  • Texas Instruments - Jai Balakrishnan, Anuj Batra,
    Anand Dabak, Srinivas Lingam
  • TDK Robert Sutton
  • University of Minnesota - Ahmed Tewfik, Ebrahim
    Saberinia, Jun Tang
  • Wisair - Gadi Shor

3
The MB-OFDM proposal is a technical merger
between
  • Texas Instruments 03/141 Batra
  • \
  • and
  • femto Devices 03/101 Cheah
  • FOCUS Enhancements 03/103 Boehlke
  • General Atomics 03/105 Ellis
  • Institute for Infocomm Research 03/107 Chin
  • Intel 03/109 Brabenac
  • Mitsubishi Electric 03/111 Molisch
  • Panasonic 03/121 Mo
  • Philips 03/125 Kerry
  • Samsung Advanced Institute of Technology
    03/135 Kwon
  • Samsung Electronics 03/133 Park
  • SONY 03/137 Fujita
  • Staccato Communications 03/099 Aiello
  • STMicroelectronics 03/139 Roberts
  • Time Domain 03/143 Kelly
  • University of Minnesota 03/147 Tewfik

4
MB-OFDM Proposal Authors
  • femto Devices J. Cheah
  • FOCUS Enhancements K. Boehlke
  • General Atomics J. Ellis, N. Askar, S. Lin, D.
    Furuno, D. Peters, G. Rogerson, M. Walker
  • Institute for Infocomm Research F. Chin,
    Madhukumar, X. Peng, Sivanand
  • Intel J. Foerster, V. Somayazulu, S. Roy, E.
    Green, K. Tinsley, C. Brabenac, D. Leeper, M. Ho
  • Mitsubishi Electric A. F. Molisch, Y.-P.
    Nakache, P. Orlik, J. Zhang
  • Panasonic S. Mo
  • Philips C. Razzell, D. Birru, B. Redman-White,
    S. Kerry
  • Samsung Advanced Institute of Technology D. H.
    Kwon, Y. S. Kim
  • Samsung Electronics M. Park
  • SONY E. Fujita, K. Watanabe, K. Tanaka, M.
    Suzuki, S. Saito, J. Iwasaki, B. Huang
  • Staccato Communications R. Aiello, T. Larsson,
    D. Meacham, L. Mucke, N. Kumar
  • STMicroelectronics D. Hélal, P. Rouzet, R.
    Cattenoz, C. Cattaneo, L. Rouault, N. Rinaldi,,
    L. Blazevic, C. Devaucelle, L. Smaïni, S.
    Chaillou
  • Texas Instruments A. Batra, J. Balakrishnan, A.
    Dabak, R. Gharpurey, J. Lin, P. Fontaine, J.-M.
    Ho, S. Lee, M. Frechette, S. March, H. Yamaguchi
  • Time Domain J. Kelly, M. Pendergrass
  • University of Minnesota A. H. Tewfik, E.
    Saberinia
  • Wisair G. Shor, Y. Knobel, D. Yaish, S.
    Goldenberg, A. Krause, E. Wineberger, R. Zack, B.
    Blumer, Z. Rubin, D. Meshulam, A. Freund

5
In addition, the following individuals/companies
support the MB-OFDM proposal
  • Fujitsu Microelectronics America, Inc A.
    Agrawal
  • Hewlett Packard M. Fidler
  • Infineon Y. Rashi
  • Jaalaa A. Anandakumar
  • Microsoft A. Hassan
  • NEC Electronics T. Saito
  • SVC Wireless A. Yang
  • TDK P. Carson
  • TRDA M. Tanahashi
  • UWB Wireless R. Caiming Qui
  • Wisme N. Y. Lee

6
Summary
  • Multiband OFDM Proposal July September
    Activities
  • Proposal Summary
  • FCC Regulation Issue
  • Time to market
  • IP statements
  • Simultaneously operation piconets (SOP)
    performance
  • Complexity and power consumption and scalability
  • CCA
  • MAC enhancements
  • Symbol Definition
  • Ranging and Location
  • Conclusions
  • QA (save your questions, reference foil number)

7
How NO votes were addressed
  • The Multi-band OFDM proposers analyzed the issues
    raised by the no-voters from the July meeting.
  • The issues were broken down into subgroups for
    technical and business evaluation
  • Each subgroup had participants from multiple
    companies
  • Held multiple meetings per week to discuss how
    best to address issues raised by no-voters and
    possible improvements
  • A number of options were evaluated and simulated
  • Summary of results and system improvements will
    be presented next
  • We are confident we have adequately addressed all
    the issues raised by the no-voters

8
Summary of Multi-band OFDM SystemPresenter
Anuj Batra (TI)
  • The complete Multi-band OFDM proposal can be
    found in the latest revision of the Word document
    03/268.

9
Overview of Multi-band OFDM
  • Basic idea divide spectrum into several 528 MHz
    bands.
  • Information is transmitted using OFDM modulation
    on each band.
  • OFDM carriers are efficiently generated using an
    128-point IFFT/FFT.
  • Internal precision requirement is reduced by
    limiting the constellation size to QPSK.
  • Information is coded across all bands in use to
    exploit frequency diversity and provide
    robustness against multi-path and interference.
  • 60.6 ns prefix provides robustness against
    multi-path even in the worst channel
    environments.
  • 9.5 ns guard interval provides sufficient time
    for switching between bands.

10
Multi-band OFDM System Parameters
  • System parameters for mandatory and optional data
    rates

Info. Data Rate 55 Mbps 80 Mbps 110 Mbps 160 Mbps 200 Mbps 320 Mbps 480 Mbps
Modulation/Constellation OFDM/QPSK OFDM/QPSK OFDM/QPSK OFDM/QPSK OFDM/QPSK OFDM/QPSK OFDM/QPSK
FFT Size 128 128 128 128 128 128 128
Coding Rate (K7) R 11/32 R 1/2 R 11/32 R 1/2 R 5/8 R 1/2 R 3/4
Spreading Rate 4 4 2 2 2 1 1
Data Tones 100 100 100 100 100 100 100
Info. Length 242.4 ns 242.4 ns 242.4 ns 242.4 ns 242.4 ns 242.4 ns 242.4 ns
Cyclic Prefix 60.6 ns 60.6 ns 60.6 ns 60.6 ns 60.6 ns 60.6 ns 60.6 ns
Guard Interval 9.5 ns 9.5 ns 9.5 ns 9.5 ns 9.5 ns 9.5 ns 9.5 ns
Symbol Length 312.5 ns 312.5 ns 312.5 ns 312.5 ns 312.5 ns 312.5 ns 312.5 ns
Channel Bit Rate 640 Mbps 640 Mbps 640 Mbps 640 Mbps 640 Mbps 640 Mbps 640 Mbps
Multi-path Tolerance 60.6 ns 60.6 ns 60.6 ns 60.6 ns 60.6 ns 60.6 ns 60.6 ns
Mandatory information data rate, Optional
information data rate
11
Band Plan (1)
  • Group the 528 MHz bands into 4 distinct groups.
  • Group A Intended for 1st generation devices (3.1
    4.9 GHz).
  • Group B Reserved for future use (4.9 6.0 GHz).
  • Group C Reserved for devices w/ improved SOP
    performance (6.0 8.1 GHz).
  • Group D Reserved for future use (8.1 10.6 GHz).

12
Multi-mode Multi-band OFDM Devices (1)
  • Having multiple groups of bands enables multiple
    modes of operations for multi-band OFDM devices.
  • Different modes for multi-band OFDM devices are
  • Future expansion into groups B and D will enable
    an increase in the number of modes.

Mode Frequency of Operation Number of Bands Mandatory / Optional
1 Bands 13 (A) 3 Mandatory
2 Bands 13, 69 (A,C) 7 Optional
13
Multi-mode Multi-band OFDM Devices (2)
  • Frequency of operation for a Mode 1 device
  • Frequency of operation for a Mode 2 device

14
FCC Compliance of MB-OFDM
  • Presenter Anand Dabak

15
Background
  • During the San Francisco IEEE meeting XSI made a
    presentation on FCC rules Slide 3 of
    03153r9P802-15_TG3a-XtremeSpectrum-CFP-Presentatio
    n.ppt
  • The issue today is NOT whether or not there is
    more or less interference
  • The issue is, what are the rules.
  • Side interest is WHY did NTIA and FCC
    specifically write rules for frequency hoppers
  • Slide 5 from 03153r9P802-15_TG3a-XtremeSpectrum-CF
    P-Presentation.ppt
  • XSI claimed that MB-OFDM needs to reduce its
    transmit power by 4.7 dB to be FCC compliant,
    MB-OFDM should transmit at -46 dBm/MHz (instead
    of -41.3 dBm/MHz)

16
Background (2)
  • MB-OFDM alliance was asked to contact the FCC to
    clarify the rules.
  • XSI/Motorola further filed a petition with the
    FCC for declaratory ruling immediately after the
    San Francisco meeting.

17
MB-OFDM Activities after S.F. to Address FCC
Issues
  • MB-OFDM alliance contacted and met with the
    FCC/OET staff, among them
  • Ed Thomas, Chief of the FCCs Office of the
    Engineering and Technology (OET)
  • Julius Knapp, Deputy Chief of the OET
  • John Reed, staff member of the OET
  • Two meetings and one phone call took place, on
    August 7th, 15th, and 23rd, 2003.

18
FCCs response
Summary of Discussions with FCC Staff Concerning
IEEE 802.15 Deliberation On Standards for
Ultrawideband devices
  • Over the past few weeks several parties have met
    with the staff of the Federal Communications
    Commission Office of Engineering and Technology
    to discuss how the Commissions rules for
    Ultrawideband devices might be applied for
    certain signal formats that are being considered
    by IEEE 802.15.
  • OET believes it is premature to make any
    determination as to the appropriate measurement
    methods for particular signals because this
    matter is under active discussion in IEEE. In
    this regard, we have no immediate plans to
    respond to the XSI/Motorola request for a
    declaratory ruling.
  • We urge that IEEE perform technical analyses to
    ensure that any UWB standard it develops will not
    cause levels of interference beyond that already
    anticipated by the rules.
  • This information will be needed to support any
    necessary FCC rules interpretations or other
    appropriate action for the chosen standard.
  • The FCC has had a long history of working
    cooperatively with the IEEE 802 committee in
    addressing any regulatory issues that may arise
    relative to standards. We recommend that IEEE
    proceed with its standards development process
    and that the committee address any questions to
    us at a later time when it has formed a specific
    proposal.

E-mail sent by Julius Knapp, Deputy Chief, OET,
FCC to XSI/Motorola and MB-OFDM proponents on
September 11th, 2003
19
FCCs response Point 1
  • FCC does not wish to become a pawn in the IEEE
    standards process, nor do they wish to stimulate
    needless new public-comment processes.
  • FCC has no current plans to respond to the
    XSI/Motorola petition for declaratory ruling.

20
FCCs response
Summary of Discussions with FCC Staff Concerning
IEEE 802.15 Deliberation On Standards for
Ultrawideband devices
  • Over the past few weeks several parties have met
    with the staff of the Federal Communications
    Commission Office of Engineering and Technology
    to discuss how the Commissions rules for
    Ultrawideband devices might be applied for
    certain signal formats that are being considered
    by IEEE 802.15.
  • OET believes it is premature to make any
    determination as to the appropriate measurement
    methods for particular signals because this
    matter is under active discussion in IEEE. In
    this regard, we have no immediate plans to
    respond to the XSI/Motorola request for a
    declaratory ruling.
  • We urge that IEEE perform technical analyses to
    ensure that any UWB standard it develops will not
    cause levels of interference beyond that already
    anticipated by the rules.
  • This information will be needed to support any
    necessary FCC rules interpretations or other
    appropriate action for the chosen standard.
  • The FCC has had a long history of working
    cooperatively with the IEEE 802 committee in
    addressing any regulatory issues that may arise
    relative to standards. We recommend that IEEE
    proceed with its standards development process
    and that the committee address any questions to
    us at a later time when it has formed a specific
    proposal.

E-mail sent by Julius Knapp, Deputy Chief, OET,
FCC to XSI/Motorola and MB-OFDM proponents on
September 11th, 2003
21
FCCs response Point 2
XSI claimed it is a rules issue
  • FCC recommends that IEEE proceed with its
    standards development process and provide
    technical analyses for OET as needed to support
    any necessary FCC rules interpretations for the
    chosen standard.
  • FCC says that a proposed UWB approach would
    likely be deemed compliant if it could be shown
    that it would not cause levels of interference
    beyond that already anticipated by the rules.
  • Contrary to XSI claims that interference is not
    the issue, FCC told MB-OFDM group that
    interference from UWB to other systems is an
    important issue.

XSI claimed it is not an interference issue
22
FCCs response
Summary of Discussions with FCC Staff Concerning
IEEE 802.15 Deliberation On Standards for
Ultrawideband devices
  • Over the past few weeks several parties have met
    with the staff of the Federal Communications
    Commission Office of Engineering and Technology
    to discuss how the Commissions rules for
    Ultrawideband devices might be applied for
    certain signal formats that are being considered
    by IEEE 802.15.
  • OET believes it is premature to make any
    determination as to the appropriate measurement
    methods for particular signals because this
    matter is under active discussion in IEEE. In
    this regard, we have no immediate plans to
    respond to the XSI/Motorola request for a
    declaratory ruling.
  • We urge that IEEE perform technical analyses to
    ensure that any UWB standard it develops will not
    cause levels of interference beyond that already
    anticipated by the rules.
  • This information will be needed to support any
    necessary FCC rules interpretations or other
    appropriate action for the chosen standard.
  • The FCC has had a long history of working
    cooperatively with the IEEE 802 committee in
    addressing any regulatory issues that may arise
    relative to standards. We recommend that IEEE
    proceed with its standards development process
    and that the committee address any questions to
    us at a later time when it has formed a specific
    proposal.

E-mail sent by Julius Knapp, Deputy Chief, OET,
FCC to XSI/Motorola and MB-OFDM proponents on
September 11th, 2003
23
FCCs response Point 3
  • FCC reminds IEEE that it has a long history of
    cooperation with IEEE on rules interpretation and
    approvals for new technologies.
  • FCC recommends that IEEE proceed with its
    standards development process and select a UWB
    proposal.

24
Summary of Presentation
  • We analyzed several different interference
    scenarios from MB-OFDM to other systems.
  • MB-OFDM does not cause any more interference than
    already anticipated by current FCC rules.
  • Hence, contrary to XSIs claims we believe
    MB-OFDM is compliant and should not have to
    reduce its transmit power by 4.7 dB to be FCC
    compliant.
  • All link budget calculations done by MB-OFDM
    group in 03267r1P802-15_TG3a-Multi-band-OFDM-CFP-P
    resentation.ppt in San Francisco are correct.
  • Lab measurements for MB-OFDM signals based upon
    MB-OFDM prototype.

25
MB-OFDM Interference Study
26
UWB Interference Considerations
  • History of FCC ruling on UWB
  • Detailed interference studies Process took 4
    years to decide with gt 1,000 comments and several
    detailed technical studies
  • Must include usage scenarios, realistic
    propagation losses, antenna patterns, etc.
  • Interference studies included considerations of
    both peak and average limits
  • Final FCC RO based upon ultra-conservative
    limits far below those placed on technologies
    that place energy into narrower portions of the
    spectrum. Separate Statement of Commissioner
    Michael J. Copps, February 14, 2002
  • Large amount of data was examined by the FCC
    before FCC ruling
  • We all have a responsibility to ensure
    interference into other radio services does not
    cause harm (example GPS, cellular and others)

27
FCC/NTIA Interference results for various US
government systems
Most systems have substantial margin available
28
FCC/NTIA Interference results for various US
government systems
Analysis based upon free-space propagation even
more conservative
DirecTV/EchoStar/DBS receivers do not operate
in 3.7-4.2 GHz they operate in the 11.7-12.2 GHz
band
29
UWB interference on Fixed Satellite Service (FSS)
Receivers
30
Average interference power is same
  • Objective is to compare MB-OFDM to impulse radios
  • FCC analyzed impulse radios before issuing RO
  • Conservative RO allows impulse radios
  • Following slides analyze UWB interference to FSS
    receivers
  • FSS frequency of operation 3.7-4.2GHz
  • DirecTV/EchoStar/DBS receivers do not operate in
    3.7-4.2 GHz they operate in the 11.7-12.2 GHz
    band
  • MB-OFDM transmitters power as described in
    MB-OFDM proposal
  • DSSS, MB-OFDM and impulse radios provide
    identical average interference power,
  • Actual performance of FSS system depends on
    specific receiver

31
UWB Waveforms for FSS Interference Analysis
DSSS
Average interference power is identical
MB-OFDM
Impulse radio
Victim receiver performance depends upon receiver
characteristic Modulation, coding, interleaving
32
  • Victim receiver Example FSS system parameters
    (implementation dependent)

FSS interference evaluation parameters
NTIA special publication 01-43, Assessment of
compatibility between ultra-wideband devices and
selected federal systems
33
Interfering Systems
  • Interfering system (1) DS-SS UWB system
  • White Gaussian noise interference
  • Interfering system (2) MB-OFDM system
    characteristics
  • Exactly as described in proposal
  • Interfering system (3) Impulse UWB radio
    characteristic
  • PRF gt 1 MHz

34
Simulation results (1)
  • 35 MSPS, rate 7/8 coding, no interleaving, Iuwb/N
    -6 dB XSI filing to FCC for typical operating
    scenarios, Sept. 2003

Very little different between UWB radios under
realistic scenarios Note SINRC/(NIsIuwb),
Issatellite intra-system interference
35
Simulation results (2)
  • 35 MSPS, rate 7/8 coding, no interleaving, Iuwb/N
    -0.4 dB

MB-OFDM interference is consistently lower than
impulse radio UWB system
36
Minimum separation between UWB radiosand FSS
receivers Example 1
  • I/N -0.4 dB, antenna elevation angle 5
    degrees, rate 7/8 coding
  • MB-OFDM transmitters can be placed closer than
    impulse radios

37
Minimum separation between UWB radiosand FSS
receivers Example 2
  • I/N -0.4 dB, antenna elevation angle 20
    degrees, rate 7/8 coding
  • MB-OFDM transmitters can be placed closer than
    impulse radios

38
Summary of FCC work
  • FCC asked MB-OFDM proponents to analyze
    interference scenarios
  • Our studies for FSS systems, show that MB-OFDM
    causes similar (almost identical) interference
    compared to other UWB systems in realistic
    scenarios.
  • MB-OFDM does not cause any interference beyond
    that which is anticipated by current FCC rules.
  • FCC said that a proposed UWB approach would
    likely be deemed compliant if it could be shown
    that it would not cause levels of interference
    beyond that already anticipated by the rules.
  • MB-OFDM proponents believe MB-OFDM is FCC
    compliant and is allowed to transmit the full
    transmit power of -41.25 dBm/MHz.
  • MB-OFDM does not have to reduce its transmit
    power by 4.7 dB as was claimed by XSI.

39
Summary of Measurement Test ResultsPresenter
Robert Sutton (TDK RD Corporation)
40
FCC Testing
  • FCC tests were performed on two different MB-OFDM
    radios (over 80 hours of lab time, more than 300
    measurements)
  • Tests performed at TDK RF Solutions EMC Test
    Services Lab
  • FCC Lab Registration No. 94066
  • NVLAP Accreditation No. 200430-0
  • Sample Measurements Performed
  • UWB Bandwidth Radiated Emissions, UWB Specific
    Requirements Emissions in GPS Bands Peak EMI
    Within a 50 MHz Bandwidth AC Mains
    Line-Conducted Disturbance Specialized
    (Conducted Antenna Terminal, Fully Anechoic)

41
Test Plan Reference
FCC CFR 47, Part 15, Subpart F Code of Federal Regulations, Part 15 Subpart F Ultra-Wideband Operation
FCC ET Docket 98-153, FCC 02-48 First RO Revision of Part 15 of the Commissions Rules Regarding Ultra-Wideband Transmissions Systems First Report Order
ANSI C63.4 1992 Methods of Measurement of Radio-Noise Emissions from Low-Voltage Electrical and Electronic Equipment in the Range of 9 kHz to 40 GHz
FCC CFR 47, Part 15, Subpart C Code of Federal Regulations, Part 15 Subpart C Intentional Radiators
FCC CFR 47, Part 15, Subpart B Code of Federal Regulations, Part 15 Subpart B Unintentional Radiators
FCC CFR 47, Part 15, Subpart A Code of Federal Regulations, Part 15 Subpart A General
CISPR 16-1 C.I.S.P.R. Specification for Radio Interference Measuring Apparatus and Measurement Methods
42
Mandatory Test Environments
1m/3m Semi-Anechoic Chamber (RE) RF Shielded
Chamber (CE)
43
Alternative Test Environments
1m/3m Fully-Anechoic Chamber (RE) Conducted
Antenna Terminal Bench (CE)
44
Device and Measurement Configuration
  • The equipment under test physical setup was done
    as prescribed in ANSI C63.4
  • The equipment under test was operating in
    accordance with its intended usage as per FCC
    2-48 First RO
  • The equipment under test was configured to
    transmit at the mandatory data rate of 110 Mbps
  • The EMI Limits were in accordance with FCC Part
    15, Subpart F

45
UWB Bandwidth and Peak Radiated Emissions within
a 50 MHz BW
Radio Sample 1 Test Distance 1m Detector
PEAK RBW/VBW 3 MHz/3 MHz Meas.
Time 1 ms Emissions lt Limit UWB
BW gt 500 MHz Note Data normalized to
3m test environment and 50 MHz RBW for limit
comparison.
46
Radiated Emissions UWB
Radio Sample 2 Test Distance 1m Detector
RMS RBW/VBW 1 MHz/3 MHz Meas. Time
1 ms Emissions lt Limit Note Data
normalized to 3m test environment for limit
comparison.
47
Emissions in GPS Bands
Radio Sample 1 Test Distance
Conducted Detector RMS RBW/VBW
1 kHz/3 kHz Meas. Time 1 ms Emissions
lt Limit Note Limit line is most stringent
at 3m distance. No emissions above noise floor
in radiated or worst case conducted measurement
mode.
48
AC Conducted Line Emissions
Radio Sample 2 Test Distance
Conducted Detector PEAK RBW/VBW
9 kHz/30 kHz Meas. Time Auto Emissions
lt Limit Note Standard AC conducted line
measurements for coupled noise.
49
Summary
  • Representative data was presented from two radio
    samples that were shown to be compliant with the
    most challenging of the UWB measurement test
    procedures
  • Additional EMI measurements (LF digital
    measurements) in accordance with FCC Part 15,
    Subpart C Intentional Radiators were performed
  • Additional EMI measurements (HF harmonic
    measurements) in accordance with FCC Part 15,
    Subpart F UWB Operation were performed
  • A series of tests in a fully anechoic chamber
    were performed to further prove compliance in an
    alternative measurement environment
  • Conducted antenna terminal tests were also
    performed
  • The EMC measurements performed under normal
    operating conditions met the FCC average and peak
    power limits

50
Time to MarketPresenter Jim Baker
(Alereon)The Claims of Rapid TTM with the XSI
Proposal You Have Heard Are at Best Misleading
51
Time to Market
  • No Voters expressed concerns about TTM
  • XSI claims much faster TTM than MB-OFDM
  • XSI claims its PHY is currently shipping
  • XSI claims to be only company with UWB silicon
  • Concerns expressed that MB-OFDM TTM would be
    unacceptable to users
  • XSI and Motorola propose a dual PHY standard and
    letting the market sort it out

All MB-OFDM Supporters are Comfortable With
MB-OFDM 1H05 TTM
52
The Truth About TTM
The PHY Work Is Not the Critical Path
  • Elements Needed For A Complete Product
  • PHY
  • MAC
  • Interoperability / Co-existence/Security
  • Association / Authentication Models
  • Applications interfaces (USB, 1394, WiMedia,
    etc)

We will work these in tandem to deliver a
COMPLETE Product in early 05
53
Approx. Timeline For A WHOLE Product
PHY / MAC
Connection Mgmt
PHY/MAC Interoperability
Convergence Arch
Applications
Time
1H 2003
2H 2003
1H 2004
2H 2004
1H 2005
2H 2005
Today, a proprietary bare pipe. A real product
that wont get returned? 05
54
Claim 1 XSI can deliver almost immediately
Response
What are the odds? 16 silicon companies with 72
voters will agree to a standard proposal
without significant change that gives one
startup a time to market advantage
This proposal will not emerge without significant
modification
55
Claim 1 Response, continued
  • XSI failed to mention
  • 1) The shipping PHY has never been described
  • and is substantially different from the current
    proposal
  • 2) Differences
  • Enhancements to the receiver for range
  • The merge elements from Parthus Ceva
  • The XSI PHY would require significant
    modifications
  • to be acceptable to the standards body - thereby
    eliminating
  • any PHY-related TTM advantage
  • The XSI PHY is not FCC certified

The current shipping design is unacceptable
56
Claim 1 Response, Cont.
  • Any company can take a non-standard or
    pre-standard product to market if
  • It passes FCC tests
  • The IP situation provides the company and its
    customers acceptable risk
  • A large portion of the UWB IP is available for
    the MB-OFDM proposal and resident within the
    companies supporting MB-OFDM
  • Working silicon is not a requirement for an IEEE
    standard as has sometimes been implied
  • CE5 did not establish early next year as their
    market horizon

57
Claim (Experience)XSI is The Only Company with
Working UWB Silicon
  • Time Domain has 2 generations of shipping
  • and FCC approved UWB silicon
  • Two Multiband OFDM compliant demos representing
  • the work of four companies are here today
  • 16 silicon companies who have a reputation for
  • shipping successful products including all
    elements
  • needed for success not just chips are
    supporters
  • of the MB-OFDM proposal

XSI has consistently overstated their ability and
experience
58
Response to Suggestion of a Dual PHY Standard
Assume for a moment that you vote to split the
standard and add dual PHYs as a necessary
condition to obtain TTM. What are the
implications?
59
Implications
The Fight Continues and Intensifies
  • Will the regulatory attacks stop?
  • XSI/Motorola have already made new interference
  • claims to the MMAC FCC
  • 2) Will WiMedia, 1394 and USB endorse one PHY or
    two?
  • Will there be one association/authentication
    model or two?
  • How will the customers know who to believe?
  • Interoperability/co-existence solutions between
    the two
  • PHYs would add significant design complexity and
    cost.
  • Manufacturers and distributors would have to
    stock multiple
  • chips and products and produce in lower volumes.

Splitting the standard makes life easy for the
IEEE and hard for everybody else, especially the
customer
60
Implications
Fragmentation Market Confusion in the Living
Room And Slow Adoption
Multiple Incompatible Radios (3) Multiple
Interoperability Forums Multiple Application
Forums Interoperability and co-existance
nightmares Multiple Association/Authentication
Schemes

Enormous Customer Confusion
And everyone can put 802.15.3a compliant on the
box
61
Implications
Delays in Reaching Critical Mass
Units
Time
The market will decide so we dont have to, but
at what cost?
62
A Better Choice
63
Work Together to Confirm and finish the Selected
Proposal
  • Support of all major CE companies
  • Support of all major silicon companies
  • Provide a complete solution, not just chips
  • Build a track record of performance, not
    undelivered claims
  • Build a cohesive market that will support all
    vendors products
  • Avoid confusing, slowing, and fragmenting the
    market
  • All the MB-OFDM companies will contribute the
    significant
  • body of IP they possess to enable the market for
    everyone
  • 8) We can still compete vigorously in the market
  • 9) Companies who can get to market fastest with
    an
  • acceptable product will be the winners and TTM
    wont be
  • damaged since the PHY is not the critical path

64
Intellectual Property Issue Presenter Roberto
Aiello (Staccato)
65
IEEE patent policy
  • Some voters have requested that we submit a
    Letter of Assurance for essential patents to the
    IEEE as condition to change their No votes
  • All author companies will conform to the IEEE
    patent policy. All companies will issue Letter of
    Assurance to the IEEE prior to the approval of
    the standard.
  • Most companies have already issued a Letter of
    Assurance to the IEEE.

66
SOP Performance Enhancement PresenterGadi Shor
(Wisair)
67
Overview
  • Goals
  • To enhance SOP performance for up to 3
    interfering piconets across the range of
    mandatory data rates.
  • Replace the baseline mode at any given data rate
    with minimal impact on other performance metrics
    (e.g., range, complexity, power consumption,
    implementation feasibility, etc).
  • July Proposal
  • Full PRF, 3- and 7-band modes with SIR
    estimation,
  • Increased interleaver depth to 6 OFDM symbols
    (600 bits),
  • 55, 110 and 200 Mbps (480 Mbps, optional).
  • Modified Proposal
  • Use Time-Spreading in place of conjugate
    symmetric spreading
  • Joint time and frequency diversity with suitable
    Interleaving Sequences
  • Coupled with increased bit-interleaver depth,
    mitigates the effect of symbol erasures due to
    collision.
  • Eliminate rotation sequences
  • Identified MAC mechanisms that can be used in
    identifying duplicate Interleaving Sequences

68
In-Band Frequency Spreading
Frequency domain spreading (frequency spreading
rate 2)
B1
A3
f3
(B1)
(A3)
A2?B2
f2
(A2?B2)
A1
B3
f1
(A1)
(B3)
t
piconetA
Collision
piconetB
Piconet A, IS f1,f2,f3,f1,f2,f3,repeat Piconet
B, IS f3,f2,f1,f3,f2,f1,repeat
Information in A2 and B2 is lost due to collision
69
Time-Spreading
  • Time domain spreading (time spreading rate 2)
  • Remove conjugate symmetric spreading in frequency
    domain
  • 200 coded bits per OFDM symbol with each symbol
    repeated in a different band according to the IS
    pattern.

B1
A2
B2
A3
f3
A1?B1
A3?B3
f2
A1
B2
A2
B3
f1
t
piconetA
Collision
piconetB
Piconet A, IS f1,f2,f3,f1,f2,f3,repeat Piconet
B, IS f3,f2,f1,f3,f2,f1,repeat
70
Interleaver Design
Puncture
G1
200 Coded bits
Interleave
Band1
QPSK mod.
IFFT
G2
Band2
200 Coded bits
200 Coded bits
Band3
IS
G3
200 coded bits per Block
200 Coded bits
convolutional encoder
200 coded bits per symbol interleaved over
3-symbol periods.
71
Interleaver Design (2)
Equivalent convolutional encoder
Example
G1
G2
G3
Combining (frequency diversity) from G2
from G3
Equivalent coding rate is 1/2
Completely erasing 200 coded bits in 600 bits due
to collision results in an equivalent coding rate
of 0.5.
72
Proposed Symbol pattern
To avoid transmitting the same symbol in the same
frequency (for achieving frequency diversity),
transmission for the time spreading is defined
as follows
IS_1
IS_2
IS_3
IS_4
73
Preliminary Results 2 SOPs
  • Preliminary SOP results with 1-interferening
    piconet in CM3 (normalized) for data rates of 110
    and 200 Mbps
  • MB-OFDM results obtained by full Monte-Carlo
    Simulation

July Proposal July Proposal Time-Spreading Time-Spreading M-BOKi M-BOKi
dint/dref 110 Mbps 200 Mbps 110 Mbps 200 Mbps 114 Mbps 200 Mbps
3-Band 0.94 1.50 0.40i 0.50 0.62 0.59(?)
7-Band 0.49 0.78 0.30i 0.30i N/A N/A
  1. Acquistion limited
  2. Low Band results from XSI WPAN Proposal, IEEE
    802.15-03/153r9, July-03

74
MAC Layer Techniques for Enhanced Piconet
Isolation
  • Scan/Remote Scan
  • Used in the event the PNC or a remote DEV detects
    the presence of a duplicate Interleaving Sequence
    (i.e. another piconet using the same channel).
  • PNC scans the environment (and/or instructs
    remote DEVs to scan) to determine which piconets
    are in use.
  • Instructs the DEVs to move to a new channel as
    needed.
  • Superframe Shuttling
  • Temporarily reduce the frame length to shift
    superframe boundaries relative to one another
    enabling one device to listen for multiple
    beacons.
  • Other
  • Child/neighbor piconets (akin to RTS/CTS??) to
    coordinate activity in proximity of a receiver
    experiencing undue adjacent piconet interference.
  • On a lightly loaded channel, determine the
    transmission pattern/periodicity in neighboring
    piconets and adjust own transmission pattern
    accordingly.

75
Complexity/Power Consumption/ ScalabilityPresente
r Charles Razzell (Philips)
76
Table of Contents
  • Multi-path energy capture using FFT vs. using
    RAKE linear equalizer
  • Digital Complexity of the major signal processing
    blocks
  • FFT
  • Viterbi
  • Discussion of 802.11a PHY scaling with process
    technology nodes
  • How big is the digital PHY in the context of the
    total system solution?
  • Scalability (in power and complexity)
  • Conclusions

77
FFT vs. Direct Linear Convolution
  • Since the UWB channel is highly dispersive,
    efficient reception requires coherently combining
    signal energy received at different times.
  • This naturally leads to a RAKE or linear FIR
    filter structure in receivers
  • It has been known since 1966 that high speed
    convolution and correlation is best performed
    using FFT/IFFT methods.
  • OFDM makes very efficient use of transforms
  • No overlap and save overheads incurred in the
    size of required FFT
  • Only one FFT block is used whether transmitting
    or receiving (dont need two transforms active
    simultaneously)
  • FFT makes a highly efficient energy collection
    engine at the heart of the MBOA proposal

78
Two Examples of Digital Energy Capture
  • RAKE receiver for DSSS (XSI)
  • Computational cost of correlators
  • Each Correlator requires 24 additions/subtractions
    at the chip-rate (e.g., 1.368GHz)
  • For RAKE of length 5, need 8 x 5 40 correlators
    for 16-BOK
  • Computational cost of MRC (this occurs at symbol
    rate. E.g., 57MHz)
  • 5 vectors of length 8 need to be multiplied by
    the corresponding 5 complex conjugate channel tap
    weights (one complex tap weight per vector).
  • 5 x 8 x 57 2280 MOPS (non-trivial complex
    multiplies)
  • Overall 2280 MOPS ( cost of 40 correlators
    running at the chip rate!)
  • FFT for MB-OFDM
  • FFT part
  • Conservatively, this requires 8 complex
    multiplies and 22.4 complex adds per clock cycle
    at 128MHz
  • Looking at complex multiplies, we need 8 x 128
    1024 MOPS
  • Frequency-domain EQ
  • 100 data carriers multiplied by their respective
    conjugate channel taps every 312.5ns
  • Implies 100/0.3125 320 MOPS
  • Overall 1344 MOPS

MB-OFDM has perfect energy capture and is
computationally less expensive.
79
XSI Multi-path Energy Capture
  • Need about 12 RAKE fingers in CM4 and 6 RAKE
    fingers in CM3 for lt 3 dB energy loss
    degradation.
  • Further performance loss can be expected due to
    inter-chip interference, when a sparse RAKE is
    used.

80
FFT Complexity (Philips Estimates)
MBOA OFDM can be achieved at very modest gate
counts for the FFT core.
81
Collated Complexity Estimates
  • Based on information in P802.15-03/213r0 and
    P802.15-03/209r3

Gates Power
FFT 70-90k (SiWorks) 55-71 mW
FFT 56k (Staccato) 54.4 mW
FFT 64k (Philips) 50.7 mW
Viterbi 75k (SiWorks) 59 mW (gt 200 Mbps) 29.7 mW ( 100 Mbps)
Viterbi 80k (Staccato) 78 mW (480 Mbps)
Viterbi lt100k (Philips) 79 mW (480 Mbps)
Multiple companies have studied the complexity of
the FFT and Viterbi cores dimensioned for the
MBOA PHY. All results point to feasible and
economic implementation in current CMOS
processes.
82
802.11a vs. MB-OFDM complexity comparison
  • Approach taken
  • Start with publicly available 802.11a
    implementation area estimates
  • Scale these numbers for CMOS process/clock
    rate/data rate etc. variations
  • Key points
  • 4 bit ADC vs. 10 bit ADC for 802.11a based on
    QPSK vs. 64-QAM, and other factors. This results
    in a scale factor of 40 for the signal
    processing area (assumed 60 of PHY) and no
    reduction for the bit processing area.
  • 132 MHz BB clock rate vs. 40MHz for 802.11a. This
    results in a scale factor of 40/132.
  • Summary of BB die area comparison results
  • MB-OFDM PHY silicon area for 110 Mbps solution
  • Approx 86 of the area of the 802.11a PHY
  • Calibration point expected area of 802.11a PHY
    in 90nm is 2.6 mm2

83
MB-OFDM vs. IEEE 802.11a power consumption
comparison
  • There are two significant components that
    differentiate the power consumption MB-OFDM
    versus IEEE 802.11a
  • Power consumption number scaled based on Intel
    ADC 5-bit, 520 MHz (linear scaling for
    frequency, doubling of power for every bit).
  • MB-OFDM consumes 500 mW less at TX and at least
    110 mW less at the receiver analog front-end when
    compared to IEEE 802.11a.

Block TFI-OFDM 802.11a
Power amp. N/A 500 mW
2 ADCs 28 mW 4-bit, 528MHz 138 mW 9-bit, 80MHz
84
Scalability
  • Data rate scaling
  • Data rates from 55 Mb/s to 480 Mb/s has been
    defined in the current proposal.
  • Frequency scaling
  • Mode 1 (3-bands) and optional Mode 2 (7-band)
    devices.
  • Guaranteed interoperability between different
    mode devices.
  • Power scaling
  • Implementers could always trade-off power
    consumption for range and information data rate.
  • Complexity scaling
  • Digital section will scale with future CMOS
    process improvements.
  • Implementers could always trade-off complexity
    for performance.

85
Power Scaling (1)
  • In a time-spreading scheme the same information
    is repeated during two symbol durations for data
    rates of 55 Mbps, 110 Mbps and 200 Mbps.
  • Power savings can be obtained by turning OFF the
    receiver during the time repeated symbols at the
    expense of system performance.
  • Majority of the digital section (except the
    Viterbi decoder) operates for only 50 of the
    time.
  • Some of the analog components (LNA, mixer and
    ADC) can be switched off for a portion of the OFF
    time. Settling/transition time is approx. 70 ns
    for ramp down and ramp up of LNA, mixer, and ADC
    in a 130 nm process ? expected off time is
    approx. 172.5 ns (27.6 of 625 ns).
  • 3 dB penalty in performance due to not collecting
    the energy from both of the time-repeated symbols.

86
Power Scaling (2)
  • Power savings estimate for the 110 Mbps data rate
    with the receiver turned OFF during the
    time-repeated symbols
  • Can achieve a power savings of 25 by paying a
    performance penalty of 3 dB.

Block High Performance Rx (ON for all time) Low Performance Rx (OFF during time-repetition)
RX AFE 121 mW 99 mW
RX Digital 84 mW 55 mW
RX Total 205 mW 154 mW
87
Complexity Scaling (1)
  • For 110 Mb/s, the MB-OFDM system has a 90 link
    success probability distance in excess of 10
    meters in all multi-path environments ? an excess
    margin of 3 dB for the 55 Mb/s mode.
  • Complexity and power consumption scaling can be
    achieved by carefully exploiting the excess
    margin. Illustrative example for 55 Mb/s
  • Can reduce both the analog and digital
    complexity/power consumption by reducing the bit
    precision
  • Ex reducing ADC precision from 4 bits ? 3 bits.
  • Ex reducing internal FFT precision (reducing the
    bit-precision of a multiplier by 30 results in a
    complexity/power consumption savings of 50).

88
Complexity Scaling (2)
  • Use of a 3-bit ADC results in 0.5 dB of
    additional degradation when compared to the use
    of a 4-bit ADC.

89
Conclusions on Scalability/Complexity
  • The heart of the Multiband OFDM proposal is an
    extremely efficient method of (de-)convolution
    based on the FFT
  • The FFT core is likely to require 64k gates or
    less at quite modest clock speeds, according to
    multiple independent studies.
  • Studies also showed that the Viterbi core for 480
    mbps is feasible in75-100k gates at the same
    clock speeds (132MHz).
  • Requiring multiple correlators (or matched
    filters) running concurrently for MBOK detection
    adds a layer of complexity to the XSI/Parthus
    proposal that we avoid.
  • These gate counts should be considered in the
    context of the overall system solution including
    hardware assisted MAC.
  • Overall, the FFT gains us excellent, robust
    performance while remaining very economic to
    implement.
  • Several options are open to implementers to
    further scale complexity and power consumption in
    exchange for performance.

90
Clear Channel Assessment for Multiband
OFDMPresenter Vern Brethour (Alereon)
91
CCA Outline
  • Overview of Clear Channel Assessment (CCA)
  • Multi-band OFDM CCA detector
  • Evaluation of XtremeSpectrum CCA
  • CCA Summary

92
Overview of CCA What is CCA?
  • Indicates co-channel DEV transmitting
  • Required for CSMA/CA to reduce collisions
  • Used w/ randomized backoff algorithm
  • Typically integrated w/ RX state machine
  • CCA is the core of contention based MACs
  • e.g., 802.11
  • In the 15.3 MAC, CCA only gates TX in CAP

93
Overview of CCA The 3 traditional flavors of CCA
  • CCA sources, usually ORed together
  • Raw Energy Detect on channel (error prone)
  • Receiving data after preamble (better)
  • Virtual Carrier Sense (best)

94
Overview of CCA Virtual Carrier Sense (best
source)
  • Calculated duration of frame
  • (Rate Length) from PHY header used
  • PHY preamble/header must be received
  • After that, oblivious to dropouts!
  • Works even if data rate not supported by
    receiver!
  • PHY header is always at base rate
  • Usage is well defined in 802.11a
  • Not yet mentioned in 802.15.3 MAC spec!

95
Overview of CCA Preamble/data received (good
source)
  • PHY preamble must be received OK
  • After that, driven by modulated symbols still
    arriving
  • Works if PHY header corrupted (preventing VCS)
  • After PHY/MAC header, only works if DEV can
    demodulate at transmitted rate
  • Functions gt RX threshold

96
Overview of CCA Energy detect (poorest source)
  • Detects anything transmitted on channel
  • Preamble can be missed (no data history)
  • Channel coherency can be lost
  • Helpful if no VCS (PHY hdr corrupt)
  • Traditionally a Close proximity CCA
  • Threshold typically (RX threshold 20dB) to
    achieve tolerable false BUSY

97
Overview of CCA Upshot on the 3 sources
  • VCS and preamble/data driven CCA
  • Any PHY with a working receiver can/should do
    both of these easily
  • Energy detect CCA
  • Very tough to do for UWB
  • Some have suggested cases for slotted Aloha
    fallback (XtremeSpectrum proposal)
  • Others have suggested preamble VCS methods are
    adequate (Sony proposal)
  • May help pathological cases, so worth
    investigating IF performance is adequate

98
Multi-Band OFDM CCA Detector Outline
  • Requirements
  • Description
  • Performance under multipath
  • Performance under MAI

99
Multi-band OFDM CCA Detector Requirements(1)
  • MB OFDM performs Preamble based CCA using
    efficient hierarchical correlator detector
  • Energy detection in channel required for
  • Energy based CCA inside a packet
  • Option for use as pre detector for the correlator
    detector
  • Detector should detect energy inside the channel
  • Channel is defined by the TFI sequence
  • Detector should be triggered only if TFI sequence
    is detected

100
Multi-band OFDM CCA Detector Requirements(2)
  • Low power consumption
  • Fast works in several usec
  • Works in multipath conditions
  • Works in the presence of frequency error between
    the transmitter and the receiver (40 ppm)
  • Works in the presence of NBI
  • Works in the presence of MAI from other piconets
  • Should have programmable threshold

101
Multi-band OFDM CCA Detector Description
  • Graph presents all three frequencies after time
    alignment according to the expected TFI sequence
  • Detects Busy only if all three are above the
    threshold at the same time
  • Reduce chances for False Alarms even when two out
    of three bands are jammed by NBI or MAI from a
    second piconet.
  • Detects energy in each sub-band robust to
    multipath and frequency errors

102
Multi-band OFDM CCA Detector Description(1)
  • Detector identifies the existence of energy in a
    given channel (TFI sequence)
  • Detector searches for energy in several
    sub-bands. Detection period is 5.6 usec.
  • The detector detects busy only if energy
    appears in all sub-bands in the correct relative
    time (based on the TFI sequence)

103
Multi-band OFDM CCA Detector Description(2)
  • The detector works in several stages to minimize
    power consumption
  • First stage detects energy in a single band
  • Second stage search for energy in other sub-bands
    based on the expected TFI sequence

104
Multi-band OFDM CCA Detector Performance CM1,
Single Piconet, SNR-5
105
Multi-band OFDM CCA Detector Performance CM2,
Single Piconet, SNR-5
106
Multi-band OFDM CCA Detector Performance CM3,
Single Piconet, SNR-5
107
Multi-band OFDM CCA Detector Performance CM1,
Two Piconets, SNR-2
108
Multi-band OFDM CCA Detector Performance CM2,
Two Piconets, SNR-2
109
Multi-band OFDM CCA Detector Performance CM3,
Two Piconets, SNR-2
110
Multi-band OFDM CCA Detector Performance CM1,
Two Piconets, SNR-1
111
Multi-band OFDM CCA Detector Performance CM2,
Two Piconets, SNR-1
112
Multi-band OFDM CCA Detector Performance CM3,
Two Piconets, SNR-1
113
Multi-band OFDM CCA Detector Summary
  • MB-OFDM proposal supports Energy based CCA
  • Detector utilize the TFI sequences to improve
    reliability
  • Detector performs well under multipath conditions
    and frequency error
  • Detector can tolerate NBI and MAI from a second
    piconet
  • Detector can detect MAI and switch to a Preamble
    based CCA

114
MB-OFDM supports CCA.What about XSIs Proposal?
  • In San Francisco, XSI showed very good results
    for CCA based on using their regular chipping
    rate as the detectable structure.
  • Different chipping rates for different channels
    allow simultaneous discrimination of all piconets
    on the air.
  • The CCA circuitry should be very power efficient.

115
Evaluation of XtremeSpectrum CCA Review of their
approach
BPF Detect
BPF
BPF Detect
LO
BPF Detect
BPF Detect
Block diagram courtesy of XtremeSpectrum, Inc.
Document 03/153 r10 dated July 2003
116
Evaluation of XtremeSpectrum CCA Simulation
Details
  • LO frequency chosen arbitrarily as 4.075GHz
  • 4.095GHz gt 20MHz, 4.101GHz gt 26MHz
  • 4.107GHz gt 32MHz, 4.113GHz gt 38MHz
  • Results are independent of LO frequency
  • After squaring circuit, 2x frequencies are
  • 40MHz, 52MHz, 64MHz, 76MHz
  • Four BP filters centered at 2x frequencies
  • 200kHz 4th order Chebshev type 1
  • Filter output squared and integrated for 5us
  • No shadowing

117
Evaluation of XtremeSpectrum CCA Free Space LO
4.095 GHzPSD of Squaring Circuit Output
Easily detectable energy in 200kHz band at IF of
40MHz.
Consistent with XtremeSpectrum results.
118
Evaluation of XtremeSpectrum CCA CM1, CIR53
piconet LO 4.095 GHzPSD of Squaring Circuit
Output
No detectable energy at the anticipated
frequency, despite high SNR (8dB).
119
What Went Wrong in Multipath?
  • x(t) squaring circuit input
  • Output y(t) x(t)2 x(t) . x(t)
  • Multiplication in the time domain is convolution
    in the frequency domain
  • X(f) dft(x(t))
  • Y(f) X(f) X(f) dft(x(t)2)
  • Y(f0) S (X(f) . X(-f f0))

120
Evaluation of XSI CCA Test Channel Used to
Explain the Failure Mechanism
20ns echo
Nulls every 50MHz
121
Evaluation of XSI CCA Fc 4.095GHz, LO
4.075GHzIF 20MHz, 2xIF 40MHz
Green curve is mirror image of blue curve,
shifted right 40MHz
Multiplying and accumulating yields the squaring
circuit output 40MHz content is easily
detectable.
122
Evaluation of XSI CCA Fc 4.101GHz, LO
4.075GHzIF 26MHz, 2xIF 52MHz
Green curve is mirror image of blue curve,
shifted right 52MHz
Multiplying and accumulating yields the squaring
circuit output 52MHz content is easily
detectable.
123
Evaluation of XSI CCA Fc 4.107GHz, LO
4.075GHzIF 32MHz, 2xIF 64MHz
Green curve is mirror image of blue curve,
shifted right 64MHz
Multiplying and accumulating yields the squaring
circuit output 64MHz content is easily
detectable.
124
Evaluation of XSI CCA Fc 4.113GHz, LO
4.075GHzIF 38MHz, 2xIF 76MHz
Green curve is mirror image of blue curve,
shifted right 76MHz
Multiplying and accumulating yields the squaring
circuit output 76MHz content is not easily
detectable.
Note the peak-to-null alignment of the 50MHz
nulls that were introduced by the channel.
125
Evaluation of XSI CCA Results in CM1,2,3,4
  • The synthetic echo test channel produces a
    scenario where a piconet with LO 4.113GHz has
    much lower probability-of-detection than do
    piconets at 4.095, 4.101, or 4.107GHz
  • Does this scenario occur in CM1,2,3,4?
  • YES

126
Evaluation of XSI CCA Average ROC Curves, CM1,
SNR 8dB
XSI result in CM1 _at_ 4m, 4.1GHz, r2 mean path
127
Evaluation of XSI CCA Average ROC Curves, CM1,
SNR 0dB
XSI result in CM1 _at_ 4m, 4.1GHz, r2 mean path
128
Evaluation of XSI CCA Average ROC Curves, CM2,
SNR 8dB
XSI result in CM2 _at_ 4m, 4.1GHz, r2.5 mean path
0.85
129
Evaluation of XSI CCA So Whats the Issue?
  • In July, XSI showed hopeful results.
  • We have independently validated their results.
  • The problem is in showing AVERAGE results.
  • Original selection criteria called out link
    probabilities of 90.
  • The CE companies want us to be showing results
    for 95.

130
Evaluation of XSI CCA 95 ROC curves, CM1, SNR
8dB
5 of CIRs in CM1 are worse than this
131
Evaluation of XSI CCA 87 ROC curves, CM1, SNR
0dB
13 of CIRs in CM1 are worse than this
132
Evaluation of XtremeSpectrum CCA Another Issue
oscillator tolerance
  • 200kHz / 4.113GHz 48.6 ppm,100kHz / 4.113GHz
    24.3 ppm,100kHz / 8.226GHz 12.15 ppm
  • Narrow bandwidth of target frequency in the
    squaring circuit output means that even modest LO
    frequency offsets cause significant CCA
    performance degradation.
  • 4 dB loss at 20 ppm (10 ppm at Tx, -10 ppm at
    Rx)
  • 13 dB loss at 40 ppm (20 ppm at Tx, -20 ppm at
    Rx)

133
Evaluation of XSI CCA Filters Used for
Detecting Coherent Energy Generated by the
Squaring Circuit
134
Evaluation of XSI CCA Situation with No LO
Error(SNR would be Slightly Better if Filter
Were Narrower)
135
Evaluation of XSI CCA Situation with /- 10ppm
LO Error(Now the Wider Filter Bandwidth is
Helping)
136
Evaluation of XSI CCA Situation with /- 20ppm
LO Error(Rapidly Falling Outside Filter Passband)
137
Evaluation of XSI CCA Average ROC Curves, CM1,
SNR 8dB/- 10ppm LO error
138
Evaluation of XSI CCA Average ROC Curves, CM2,
SNR 8dB/- 10ppm LO error
139
Evaluation of XSI CCA 90 ROC curves, CM1, 8
dB, /- 10ppm
10 of CIRs in CM1 are worse than this
140
Evaluation of XSI CCA Average ROC Curves, CM1,
SNR 8dB/- 20ppm LO error
141
Evaluation of XSI CCA 75 ROC curves, CM1, 8
dB, /- 20ppm
25 of CIRs in CM1 are worse than this
142
Evaluation of XSI CCA Average ROC Curves, CM2,
SNR 8dB/- 20ppm LO error
143
Summary of XSI Squaring Circuit CCA Mechanism
  • Works well under ideal conditions
  • Suffers in the presence of multipath
  • Suffers in the presence of LO mismatch
  • Many channels fail even at high SNR
  • Performance problems demonstrated in CM1 for the
    low frequency band
  • Worse in CM2, CM3, CM4
  • Worse in the high (8.2GHz) frequency band
  • Not a reliable CCA mechanism

144
CCA Summary
  • The best CCA forms are VCS and preamble/data
  • Energy detect driven CCA is nice addition, if a
    UWB solution possible
  • There is a good CCA solution for MB OFDM!
  • Performs well in multipath
  • Is challenged by SOP, but this is detectable w/
    fallback to preamble/VCS methods
  • The XtremeSpectrum CCA is not reliable
  • Very challenged in multipath (CM2-gtCM4), and even
    in many CM1 channel realizations
  • No practical detection / fallback methods

145
MAC Enhancements for Multiband OFDMPresenter
Larry Taylor (Staccato Communications)
146
MAC Enhancement Outline
  • Transparency to 15.3
  • Value-add enhancements
  • Band management / frequency agility
  • Ranging support
  • MAC Enhancement Conclusions

147
Transparency to 15.3 Overview
  • Multi-band OFDM PHY has default modes
  • Band management is part of an optional value-add
    story
About PowerShow.com