Doc.: IEEE 802.11-04-0020-00-000n Submission Jan 2004 M.Faulkner, ATcrcSlide 1 Low Overhead Pilot Structures Igor Tolochko and Mike Faulkner, ATcrc, Victoria.

Slides:



Advertisements
Similar presentations
Doc.: IEEE /0071r1 Submission January 2004 Aleksandar Purkovic, Nortel NetworksSlide 1 LDPC vs. Convolutional Codes for n Applications:
Advertisements

Doc.: IEEE r0 Submission November 2002 Je Woo Kim, TeleCIS WirelessSlide 1 PAPR Reduction of OFDM by Unitary Transformations Je Woo Kim TeleCIS.
Doc.: IEEE /1212r0 Submission October Jeng-Hong Chen, Pansop Kim, Winbond ElectronicsSlide 1 Performance of Circulation Transmission (Sub_BC)
Prakshep Mehta ( ) Guided By: Prof. R.K. Shevgaonkar
Anatomy of Radio LAN Onno W. Purbo
Doc.: IEEE / k Submission September 2003 Brian Johnson, Nortel Networks a Performance Over Various Channels 9 September 2003 Presenter:
Semi-Blind Equalization for OFDM using Space-Time Block Coding and Channel Shortening Alvin Leung Yang You EE381K-12 March 04, 2008.
Doc.: IEEE /0630r0 Submission May 2015 Intel CorporationSlide 1 Verification of IEEE ad Channel Model for Enterprise Cubical Environment.
Doc.: IEEE /xxxx Submission Control PHY Design for 40-50GHz Millimeter Wave Communication Systems.pptx Authors: May 2015 Slide 1Jianhan Liu, et.
1 Channel Estimation for IEEE a OFDM Downlink Transmission Student: 王依翎 Advisor: Dr. David W. Lin Advisor: Dr. David W. Lin 2006/02/23.
MIMO-OFDM MIMO MIMO High diversity gain (space-time coding) High diversity gain (space-time coding) High multiplexing gain (BLAST) High multiplexing gain.
Muhammad Imadur Rahman1, Klaus Witrisal2,
12- OFDM with Multiple Antennas. Multiple Antenna Systems (MIMO) TX RX Transmit Antennas Receive Antennas Different paths Two cases: 1.Array Gain: if.
Doc.: IEEE /1391r0 Submission Nov Yakun Sun, et. Al.Slide 1 About SINR conversion for PHY Abstraction Date: Authors:
Doc.: IEEE /0489r1 Submission May 2010 Alexander Maltsev, IntelSlide 1 PHY Performance Evaluation with 60 GHz WLAN Channel Models Date:
Doc.: IEEE /1399r0 Submission November 2014 Multi-Carrier Training Field for OFDM Transmission in aj (45GHz) Authors/contributors: Date:
Doc.: IEEE /0705r1 Submission Control PHY Design for 40-50GHz Millimeter Wave Communication Systems Authors: May 2015 Slide 1Jianhan Liu, et.
Doc.: IEEE /180r0 Submission March 2002 Monisha Ghosh, et al., Philips Slide 1 On The Use Of Multiple Antennae For Monisha Ghosh, Xuemei.
Submission doc.: IEEE 11-13/0536r0 May 2013 Wookbong Lee, LG ElectronicsSlide 1 HEW SG PHY Considerations For Outdoor Environment Date: Authors:
1 PERFORMANCE OF FREQUENCY OFFSET SYNCHRONIZATION IN A SINGLE AND MULTI-ANTENNA IEEE SYSTEM José A. Rivas Cantero M. Julia Fernández-Getino.
Doc.: IEEE /0075r0 Submission January 2004 H. Sampath, PhD, Marvell SemiconductorSlide 1 Pros and Cons of Circular Delay Diversity Scheme for.
Doc.: n-proposal-statistical-channel-error-model.ppt Submission Jan 2004 UCLA - STMicroelectronics, Inc.Slide 1 Proposal for Statistical.
NTUEE Confidential Toward MIMO MC-CDMA Speaker : Pei-Yun Tsai Advisor : Tzi-Dar Chiueh 2004/10/25.
Space-Time and Space-Frequency Coded Orthogonal Frequency Division Multiplexing Transmitter Diversity Techniques King F. Lee.
| Anja Sohl Pilot Assisted and Semiblind Channel Estimation for Interleaved and Block-Interleaved Frequency Division Multiple Access Anja Sohl.
Doc.: IEEE /1401r0 Submission November 2014 Slide 1 Shiwen He , Haiming Wang Quasi-Orthogonal STBC for SC-PHY in IEEE aj (45GHz) Authors/contributors:
Doc.: IEEE /1645r2 Submission January 2005 C. Hansen, BroadcomSlide 1 Preambles, Beamforming, and the WWiSE Proposal Notice: This document has.
Doc.: IEEE /232r0 Submission March 2002 Todor Cooklev, AwareSlide 1 Extended Data Rate a Marcos Tzannes Todor Cooklev Dongjun Lee Colin.
Improved Channel Estimation Based on Parametric Channel Approximation Modeling for OFDM Systems IEEE TRANSATIONS ON BROADCASTING , VOL. 54 NO. 2 JUNE.
Doc.: IEEE /0161r1 Submission doc.: IEEE /0087r0 January 2010 R. Kudo, K. Ishihara and Y. Takatori (NTT) Slide 1 Measured Channel Variation.
Doc.: IEEE /0075r1 Submission January 2004 H. Sampath, R. Narasimhan, Marvell SemiconductorSlide 1 Advantages and Drawbacks of Circular Delay.
A Simple Transmit Diversity Technique for Wireless Communications -M
Doc.: IEEE /1229r1 Submission November 2009 Alexander Maltsev, IntelSlide 1 Application of 60 GHz Channel Models for Comparison of TGad Proposals.
Doc.: IEEE /0929r1 Submission August 2004 Patrik Eriksson et. al., WaveBreaker ABSlide 1 A “High Throughput” Partial Proposal Patrik Eriksson,
Doc.: IEEE /0929r0 Submission August 2004 Patrik Eriksson et. al., WaveBreaker ABSlide 1 A “High Throughput” Partial Proposal Patrik Eriksson,
Doc.: IEEE /0632r0 Submission May 2008 Vinko Erceg, BroadcomSlide 1 VHT 60 GHz Channel Model Recommendation Date: Authors:
Doc.: IEEE /0779r0 Submission Guixia Kang, BUPT July 2010 VHT-LTF Design for IEEE802.11ac Slide 1 Date: Authors:
Doc.: IEEE /159r0 Submission March 2002 S. Hori, Y Inoue, T. Sakata, M. Morikura / NTT. Slide 1 System capacity and cell radius comparison with.
Doc.: IEEE Submission Nov 2014 Link Budget Analysis for 40-50GHz Indoor Usage Date: Authors: Slide 1Jianhan Liu, Mediatek Inc.
Doc.: IEEE /0929r2 Submission September 2004 Patrik Eriksson et. al., WaveBreaker ABSlide 1 A “High Throughput” Partial Proposal Patrik Eriksson,
Doc.: IEEE /553r0 Submission May 2004 Ravi Mahadevappa, Stephan ten Brink, Realtek Slide 1 MIMO Mode Table for n Ravi Mahadevappa,
Performance Analysis of Open Loop SU-MIMO Receivers for IEEE ay
Doc.: IEEE /0174r1 Submission February 2004 John Ketchum, et al, QualcommSlide 1 PHY Abstraction for System Simulation John Ketchum, Bjorn Bjerke,
Introduction to OFDM and Cyclic prefix
Proposal for Statistical Channel Error Model
Space-Time and Space-Frequency Coded Orthogonal Frequency Division Multiplexing Transmitter Diversity Techniques King F. Lee.
Satoru Hori, Yasuhiko Inoue, Tetsu Sakata, Masahiro Morikura
Effective (20us) Preambles for MIMO-OFDM
Diversity Lecture 7.
Space Time Coding and Channel Estimation
Transmit diversity for MIMO-OFDM
Linglong Dai and Zhaocheng Wang Tsinghua University, Beijing, China
Tsinghua University, Beijing, China
Validation of n Channel Models
A Novel Multiple Access System Based on TDS-OFDM
802.11n Channel Model Validation
Coherence Time Measurement in NTT Lab.
6-10GHz Rate-Range and Link Budget
HDR a solution using MIMO-OFDM
Backwards compatibility
Multicarrier Communication and Cognitive Radio
Enhanced Beam Tracking Against Blockage: Resolution to CID 145
Multi-band Modulation, Coding, and Medium Access Control
Multi-band Modulation, Coding, and Medium Access Control
Preambles for MIMO channel estimation
STBC in Single Carrier(SC) for IEEE aj (45GHz)
20us effective preambles for MIMO-OFDM
20us effective preambles for MIMO-OFDM
PHY Performance Evaluation with 60 GHz WLAN Channel Models
Enhanced Beam Tracking Against Blockage: Resolution to CID 145
Presentation transcript:

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 1 Low Overhead Pilot Structures Igor Tolochko and Mike Faulkner, ATcrc, Victoria University

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 2 MIMO System Channel Estimatiom –Mt * Mr paths –S is the total transmit power: Tx power / antenna = S/Mt (equal power) –SNR is the receiver signal to noise ratio –MSE is the channel estimate error Goal Estimate all Paths With the Minimum Overhead PA 1 PA 2 PA 3 PA Mt RX 1 RX 2 RX 3 RX Mr AMt A2 A3 A1

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 3 Channel Estimation Requirements Mt x Mr MIMO requires Mt orthogonal pilots –Time orthogonality –Frequency –Code Low Pilot Overhead Low Complexity Low implementation (SNR) loss (low mean squared error) –MSE(channel) << SNR (data) –MSE < SNR -3dB (for 11a) SNR loss <= 1.76dB Channel MSE SNR Loss Comment SNR - inf(dB) 0dBPerfect Channel Estimation SNR – 6dB 0.97dB SNR – 3dB 1.76dBLong Symbol + Least Squares SNR3dBUse Differential Detection?? SISO Coherent Detection

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 4 Time Orthogonal One Transmitter at a time –Each PA must provide the full power, S, during pilot transmission –Overhead 2Mt symbols per packet –Simple LS processing for MSE=0.5N (-3dB) –11a Compatibility by inserting the ‘signal’ symbol after the first long, P(1), and transmitting it on A1 as per Jan Boer et al [2] 8  s Sub-carriers A1 A2 AMt 8  s Time Interleaved Long (TIL) Pilots P(1) P(2) P(Mt) Time Frequency cp

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 5 Code Orthogonality All Tx at same time –P(k)= k’th pilot –PA power of S/Mt during pilot transmission –Overhead 2Mt symbols per packet –Simple LS processing for MSE=0.5N (-3dB) –11a compatibility by inserting the ‘signal’ symbol after the first long symbol P(1) and transmitting it using co-transmission on A1 and A2 8  s Sub-carriers A1+A2 A1-A2 Example Mt=2 Walsh like Can be expanded to Mt antennas (low detection complexity if Mt=2,4,8) Code Interleaved Long (CIL) Pilots P(1) P(2) 8  s cp

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 6 Frequency Orthogonality All Tx active, but not in the same frequency bin –PA power of S/Mt during pilot transmission –Overhead of 2 symbols (1 long) per packet … but MSE=0.5N + interpolation error …. too high?? –Might need Longer Pilots, More Pilots or Frequency Domain Processing to reduce MSE. Simulations of Interpolation performance needed 8  s Sub-carriers A1 A2 Receiver uses Interpolation for non pilot positions AMt A1 Frequency Interleaved Long (FIL) Pilots cp

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 7 8  s 4  s 8  s Sub-carriers (a)(c) A1 A2 A1 A2 A1 A2 A1 A2 A1 A2 A1 A2 A1 A2 A1 A2 A1 A2 A1 (b) A1 A2 A1 A2 A1 A2 A1 A2 A1 A2 The pilot schemes in a 2x1 OFDM system: (a) Frequency Interleaved Long (FIL). (b) Frequency and Time Interleaved Normal (FTIN) and (c) Double Frequency Interleaved Long (DFIL). Low Overhead Pilot Symbol Schemes to be Studied Note. FTIN (b) and DFIL (c) are not compatible to 11a as a first pilot signal FIL FTIN DFIL cp 4  s cp

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 8 Frequency Domain Channel Estimation [5] LS estimation is a noisy observation, given by: Linear minimum mean-squared error (LMMSE) estimator is a conditional mean estimator: LMMSE exploits frequency domain correlation of h(t) and performs interpolation and filtering of the noisy LS estimation Close to optimal (Weiner Filter) Assumptions Perfect synchronisation and SNR knowledge Exponential decaying channel impulse response LMMSE Processing + H

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 9 2x1 diversity, IEEE n NLOS channels [1] Channel B,C,D,  rms <= 50 ns Channel Estimate MSE vs SNR Channel E,  rms = 100 ns data SNR(dB) Mean Square Error(dB) LS in TIL or CIL LMMSE in FIL LMMSE in FTIN LMMSE in DFIL FTIN is the best of the single long pilot structures LMMSE smoothing/interpolation is best at low SNR

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 10 Channel F,  rms = 150 ns 2x1 diversity, IEEE n NLOS channels Channel Estimate MSE vs SNR Typical Floor in Data that might be expected by ISI due to channel and impulse response Tx/Rx filters. Data is corrupted more than the Channel Estimate Channel F + filter ISI

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 11 Channel E,  rms = 100 ns. Reduced complexity (RC) LMMSE Channel F,  rms = 150 ns. Reduced complexity (RC) LMMSE 2x1 diversity, IEEE n NLOS channels

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 12 Channel E,  rms = 100 ns. Reduced complexity (RC) LMMSE Channel F,  rms = 150 ns. Reduced complexity (RC) LMMSE 3x1 diversity, IEEE n NLOS channels Performance Criteria: SNRmax (when MSE=SNR-3dB) :- the higher the better SNRmax

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 13 Table I. Single Pilot Symbol Overhead = 8us (1 long or 2 normal) Scheme SNRmax (dB) when MSE=SNR-3dB 21214141 FIL FTIN DFIL FIL/RC > 37 > 37 > – – 313 – – EE E F FF Table II. Double Pilot Symbol Overhead = 16us (2 long) Scheme SNRmax (dB) when MSE=SNR-3dB 21214141 FTIL > 37 > 131 – – EE E F FF SNR range for LMMSE to outperform Time Interleaved LS Figures over 20dB are probably acceptable Add between 4 to 10 dB to get the condition MSE =SNR

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 14 Conclusion Frequency domain processing can reduce pilot overhead to Mt normal symbols (Mt * 4  s). Example 8  s for 2x1 Interpolation error increases with delay spread Low overhead pilots perform best at Low SNR. Limiting performance in high delay spread channels with high SNR (are these common?) FTIN has reduced interpolation error and outperforms FIL, except in ISI channels. It is the best of the low overhead schemes. Frequency interleaving (FIL) is preferred because of compatibility. ISI corrupts data more than the channel MSE COMPLEXITY is HIGH –Complexity/bin Cp = No of MIMO Channels * no of operations per bin –For full LMMSE, Cp = (Mr*Mt)*(N/Mt) operations/bin Cp = Mr*N ; N=no. of active bins (=52) –RC systems have between x0.3 to x0.8 reduction.

doc.: IEEE n Submission Jan 2004 M.Faulkner, ATcrcSlide 15 Reference 1. V. Erceg et al, TGn Channel Models, doc.: /940r, Nov Jan Boer et al. ‘ Backwards compatibility’,doc.: IEEE /714r0, Sept IEEE Std a/D7.0, “Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications: High-Speed Physical Layer in the 5 GHz Band,” New York, USA, J. Medbo and P. Schramm, “Channel Models for HIPERLAN/2 in Different Indoor Scenarios”, ETSI BRAN doc. No. 3ER1085B, O. Edfors, M. Sandell, J.J. van de Beek, S.K. Wilson and P.O. Borjeson, “OFDM Channel Estimation by Singular Value Decomposition,” IEEE Trans. Commun., 46,(7), pp. 931–939, I. Tolochko, and M. Faulkner, “Real Time LMMSE Channel Estimation for Wireless OFDM Systems with Transmitter Diversity,” In Proc. IEEE 56th Vehicular Technology Conference VTC2002-Fall, Vancouver, Canada, Sept. 2002, pp. 1555–1559.