CSI Feedback for MIMO-OFDM Transmission in IEEE aj (45 GHz)

Slides:



Advertisements
Similar presentations
Doc.:IEEE /1275r0 Submission Laurent Cariou Nov, 2010 Slide 1 Complexity reduction for time domain H matrix feedback Authors: Date:
Advertisements

Cyclic Shift Diversity Design for IEEE aj (45GHz)
Doc.: IEEE /1196r1 Submission Data Rate and Spectrum Requirements for IEEE aj (45 GHz) Date: Authors: Haiming Wang (SEU)Slide.
Doc.: IEEE /1090/r2 Submission September 2013 Submission Zhanji Wu, et. Al. Non-linear pre-coding MIMO scheme for next generation WLAN Date:
Doc.: IEEE /1539r0 Submission Dec Minho Cheong, ETRISlide 1 Beam forming for 11ah Date: Authors:
Doc.: IEEE /xxxx Submission Control PHY Design for 40-50GHz Millimeter Wave Communication Systems.pptx Authors: May 2015 Slide 1Jianhan Liu, et.
May 2015 Submission doc.: aj Shiwen He, Haiming Wang PPDU Format for IEEE aj (45GHz) Authors/contributors: Date: Presenter:
May 2015 Submission doc.: aj Shiwen He, Haiming Wang PPDU Format for IEEE aj (45GHz) Authors/contributors: Date: Presenter:
May 2015 Submission doc.: aj Shiwen He, Haiming Wang PPDU Format for IEEE aj (45GHz) Authors/contributors: Date: Presenter:
Doc.: IEEE /0716r6 Submission PHY SIG Frame Structure for IEEE aj (45GHz) Authors/contributors: Date: Presenter: Shiwen HE May 2015.
Doc.:IEEE /0206r0 Submission January 2015 Shiwen He, Haiming Wang Pilot Design for OFDM PHY for aj(45 GHz) Authors/contributors: 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 /0364r1 SubmissionEldad Perahia, Intel CorporationSlide 1 Date: Authors: Antenna Array Gain from Measured Data for n/ac.
Doc.: IEEE /0161r1 Submission doc.: IEEE /1131r0 Sept K. Ishihara et al.,(NTT) Slide 1 Sept Slide 1 Time-Domain CSI Compression.
Doc.: IEEE /0831r0 Submission July 2010 Yusuke Asai (NTT)Slide 1 Frame Sequence of Interference Management Using Beamforming Technique in OBSS.
Doc.: IEEE /1401r0 Submission November 2014 Slide 1 Shiwen He , Haiming Wang Quasi-Orthogonal STBC for SC-PHY in IEEE aj (45GHz) Authors/contributors:
Submission doc.: IEEE /1088r0 September 2015 Daewon Lee, NewracomSlide 1 LTF Design for Uplink MU-MIMO Date: Authors:
Doc.: IEEE /314r0 Submission March 2004 Taehyun Jeon, ETRISlide 1 Adaptive Modulation for MIMO-OFDM Systems Taehyun Jeon, Heejung Yu, and Sok-kyu.
Submission September 2015 doc.: IEEE /1091r0 September 2015 Considerations on Range Extension with SIG-A Repetition Date: Authors:
Doc.: IEEE /0112r0 Zhanji Wu, et. Al. January 2013 Submission Joint Coding and Modulation Diversity for the Next Generation WLAN Date:
Submission doc.: IEEE /1347r0 November 2015 Filippo Tosato, ToshibaSlide 1 Strategies to reduce MIMO feedback overhead Date: Authors:
Doc.: IEEE /1398r0 Submission November 2014 Slide 1 Shiwen He, Haiming Wang Preamble Sequence for IEEE aj (45GHz) Authors/contributors:
Doc.: IEEE /0883r0 Submission PHY SIG Frame Structure for IEEE aj (45GHz) Authors/contributors: Date: February 5, 2016 Presenter: Haiming.
Doc.: IEEE aj Submission PHY SIG Frame Structure for IEEE aj (45GHz) Authors/contributors: Date: Presenter: Shiwen.
Doc.: aj Submission November 2014 Slide 1 Shiwen He , Haiming Wang Quasi-Orthogonal STBC for IEEE aj ( 45GHz ) Authors/contributors:
Doc.: IEEE /1321r0 Submission Reducing Explicit MIMO Compressed Beamforming Feedback Overhead for ax November 2015 Slide 1 Date:
Submission doc: IEEE /0807r0 July 2010 R. Kudo et al., NTT Slide 1 PHY Abstraction for MU-MIMO Date: Authors: Name AffiliationsAddressPhone .
Doc.:IEEE /1227r0 Submission Nov, 2010 Joonsuk Kim, et al Slide 1 11ac Explicit Feedback Format Authors: Date:
Doc.: IEEE /0883r2 Submission September 2014 PHY SIG Frame Structure for IEEE aj (45GHz) Authors/contributors: Date: Presenter:
Doc.: IEEE /0205r0 Submission Jan 2015 Shiwen He, Haiming Wang Slide 1 Time Domain Multiplexed Pilots Design for IEEE802.11aj(45 GHz) SC PHY Authors/contributors:
Doc.: IEEE /1499r0 Submission September 2006 Andrew Storm, BandspeedSlide 1 Enhancement of the Compressed steering matrix performance with power.
Doc.: IEEE /0161r1 Submission doc.: IEEE /0806r0 K. Ishihara et al.,(NTT) Slide 1 July 2010 Slide 1 CSI Feedback Scheme using DCT for.
Doc.:IEEE /1105r0 Submission Sept 14, 2010 Hongyuan Zhang, et al Slide 1 11ac Explicit Sounding and Feedback Authors: Date:
Doc.: IEEE aj Submission PHY SIG Frame Structure for IEEE aj (45GHz) Authors/contributors: Date: Presenter: Haiming.
Doc.: IEEE /0935r1 Submission July 2011 Fei Tong, CSRSlide 1 An improved non-compressed beamforming feedback format for 11ac Date:
Doc.: IEEE /0626r1 Submission Feedback Element Compression for ax May 2016 Slide 1 Date: Authors: Kome Oteri (InterDigital)
Closed Loop SU-MIMO Performance with Quantized Feedback
Length 1344 LDPC codes for 11ay
Discussions on 11ac PHY Efficiency
Comparisons of Simultaneous Downlink Transmissions
Maximum Tone Grouping Size for ax Feedback
Feedback Element Compression for ax
Hybrid Beamforming Protocol Design Details
Maximum Tone Grouping Size for ax Feedback
Feedback Element Compression for ax
Maximum Tone Grouping Size for ax Feedback
Discussions on 11ac PHY Efficiency
Preamble Sequence for aj(45GHz)
Further Discussion on Beam Tracking for ay
PHY SIG Frame Structure for IEEE aj (45GHz)
Channel Dimension Reduction in MU Operation
Hybrid Beamforming Protocol Design Details
Time Domain CSI report for explicit feedback
802.11ac Preamble Date: Authors: Month Year Month Year
Reducing Channel Dimension in MU-MIMO CSI Feedback
PPDU Format for IEEE aj (45GHz)
Discussions on 11ac PHY Efficiency
Discussions on 11ac PHY Efficiency
11ac Explicit Sounding and Feedback
Time Domain CSI report for explicit feedback
CSI Feedback Scheme using DCT for Explicit Beamforming
STBC in Single Carrier(SC) for IEEE aj (45GHz)
Reducing Channel Dimension in MU-MIMO Explicit Feedback Operation
MU-MIMO Explicit Feedback Dimension Reduction Procedures
An improved non-compressed beamforming feedback format for 11ac
Strawmodel ac Specification Framework
Symbol Interleaving for Single Carrier PHY in aj (45 GHz)
LDPC Tone Mapping for IEEE aj(45GHz)
Considerations on feedback overhead
Presentation transcript:

CSI Feedback for MIMO-OFDM Transmission in IEEE 802.11aj (45 GHz) Date: 2014-11-5 Presenter: Haiming Wang Authors/contributors:

Abstract This presentation proposes CSI feedback schemes for transmit beamforming in IEEE 802.11aj (45 GHz).

Introduction (1/2) Beamforming can improve the performance of system, including Enhance throughput in IEEE 802.11n/ac Quasi-ML detection performance can be achieved with a low-complexity receiving structure. Expand coverage in IEEE 802.11ad By focusing transmitting power on a specific direction, signals can be transmitted to a longer distance. Compressed beamforming matrix feedback based on Givens Rotation has been used in IEEE 802.11n/ac, due to Reduced feedback overhead Low complexity

Introduction (2/2) The number of bits used for angle quantization in IEEE 802.11n and IEEE 802.11ac 802.11n supports (3,1), (4,2), (5,3) or (6,4) bits to quantize angle (ϕ, ψ). 802.11ac supports (4,2) or (6,4) bits to quantize angle (ϕ, ψ) for single user, and (7,5) or (9,7) bits to quantize angle (ϕ, ψ) for multi-user. Subcarrier grouping has been applied in IEEE 802.11n and IEEE 802.11ac to further reduce feedback amount 11n and 11ac both support to combine 2 or 4 subcarriers into one group. Appropriate interpolation method is needed to reconstruct beamforming matrices.

Feedback scheme Explicit feedback is proposed for beamforming to 802.11aj (45 GHz), including CSI feedback Channel matrix H Noncompressed Beamforming Matrix feedback Right singular matrix of H Compressed Beamforming Matrix feedback Compressed right singular matrix of H

Angle Quantization For compressed beamforming matrix feedback based on Givens Rotation , angles ψ and ϕ are quantized as where and are the number of bits used to quantize ψ and ϕ respectively. After quantization, angle ϕ is quantized between 0 and 2π, angle ψ is quantized between 0 and π/2. is more than by 2 bits.

Subcarrier Grouping For subcarrier grouping, the group size should satisfy where is subcarrier frequency spacing, is the coherent bandwidth of the channel. The RMS delay spread of 802.11aj (45 GHz) channel is 10 ns, and , , . Since the number of effective subcarriers is 176/352, which is even, so optional set is {2, 4, 6}.

Frame Format of NDP Propose to use the same NDP sounding mechanism as 11ac, and the NDP format is shown as follows. QTF is composed of 14 ZCZ sequences MCTF is used to estimate channel, and N depends on the dimension of channel matrices to be estimated.

Frame Format of NDP QMG NDP Announcement frame format Feedback Type Set to 0 for SU; Set to 1 for MU. Nc Index If the Feedback Type field indicate MU, then Nc Index indicates the number of columns Nc of feedback matrix: Set to 0 \1\2\3 to request Nc = 1\2\3\4 Reserved if the Feedback Type field indicates SU.

Frame Format of MIMO Control QMG CSI/Beamforming frame format QMG MIMO Control field Order Information 1 Category 2 QMG Action 3 QMG MIMO Control 4 QMG CSI/Beamforming Report 5 MU Exclusive Noncompressed /Compressed Beamforming Report The Category field is set to 22 for QMG Action The QMG Action field is set to 0 for QMG CSI, set to 1 for QMG Noncompressed Beamforming, set to 2 for QMG Compressed Beamforming. The MU Exclusive Noncompressed/Compressed Beamforming Report present when the Feedback Type is MU.

Description of MIMO Control Field QMG MIMO Control field description Nc Index Indicates the number of columns of V matrix: Set to 0\1\2\3 for Nc=1\2\3\4 Nr Index Indicates the number of rows of V matrix: Set to 0\1\2\3 for Nr=1\2\3\4 Channel Width Indicates the channel width: Set to 0\1 for 540\1080 MHz Grouping Indicates the number of carriers grouped into one: Set to 0\1\2\3 for Ng=1\2\4\6 Codebook Information Indicates the number of bits in the representation of the real and imaginary parts of each element in the matrix for QMG CSI feedback and QMG Noncompressed Beamforming feedback, or indicates the size of codebook entries for Compressed Beamforming feedback: For CSI feedback: Set to 0\1\2\3 for Nb = 4\5\6\8 For Noncompressed Beamforming feedback: Set 0\1\2\3 for Nb = 4\3\6\8 For compressed Beamforming feedback: If Feedback Type is SU: If Feedback Type is MU: Set to 0 for 2 bits for ψ, 4 bits for ϕ Set to 0 for 5 bits for ψ, 7 bits for ϕ Set to 1 for 3 bits for ψ, 5 bits for ϕ Set to 1 for 7 bits for ψ, 9 bits for ϕ

Simulation Settings Channel model: 802.11aj (45 GHz) channel Number of distinguishable paths: 25 Maximum/RMS delay spread: 100 ns/10 ns Channel bandwidth: 540 MHz Packet length: 4096 bytes Number of channel realizations: 3000 Simulation antennas: 2×1, 4×1 for 1ss, 3×2, 4×2, 4×4 for 2ss, 4×3 for 3ss. Modulation and code rate: {QPSK ½},{64QAM ⅝} Single user, LS channel estimation, without STBC. Actual channel estimation for receiving sounding NDP is added. Linear spherical interpolation is applied for subcarrier grouping, and use 7 bits to quantize ϕ, 5 bits to quantize ψ.

Simulation Results For Givens Rotation based angle quantization, simulation show that Using 5 bits to quantize ϕ, 3 bits to quantize ψ could achieve the performance of perfect beamforming matrix. Using 4 bits to quantize ϕ, 2 bits to quantize ψ could also achieve the performance close to perfect beamforming matrix, with performance loss less than 0.4 dB. For subcarrier grouping, simulations show that For , the maximum performance loss is 1.8 dB For , the maximum performance loss is 2 dB For , the maximum performance loss is 2.5 dB For , the maximum performance loss is 3.4 dB

Conclusions Two type of angle quantization are proposed to IEEE 802.11aj (45 GHz) , including 4 bits to quantize ϕ, 2 bits to quantize ψ . 5 bits to quantize ϕ, 3 bits to quantize ψ . Optional group size set {1, 2, 4, 6} is proposed for subcarrier grouping in IEEE 802.11aj (45 GHz).

Reference [1] “11-10-0332-00-00ac-csi-report-for-explicit-feedback-beamforming-in-downlink-mu-mimo”, Koichi Ishihara et al. [2] “11-10-0806-01-00ac-csi-feedback-scheme-using-dct-for-explicit-beamforming”, Koichi Ishihara et al. [3] “11-11-1539-00-00ah-beamforming-for-11ah”,Minho Cheong et al. [4]“11-05-1645-02-000n-preambles-beamforming-wwise-proposal”,Christopher J. Hansen et al. [5]"11-07-0612-02-000n-comment-resolution-csi-uncompressed-steering-matrix-feedback-bitwidth-nb",Hongyuan Zhang et al. [6]"11-10-0586-01-00ac-time-domain-csi-report-for-explicit-feedback ", Laurent Cariou et al. [7]"11-10-1131-00-00ac-time-domain-csi-compression-schemes-for-explicit-beamforming-in-mu-mimo",Koichi Ishihara et al. [8]"Draft P802.11REVmc_D1.5" [9]"Draft-802.11ac_D5.1"

Simulation Results for Angle Quantization APPENDIX A: Simulation Results for Angle Quantization

can achieve performance close to unquantized angles, with 0.1 dB performance loss.

can achieve performance close to unquantized angles, with 0.1 dB performance loss.

can achieve performance close to unquantized angles, with 0.1 dB performance loss.

can achieve performance close to unquantized angles, with 0.2 dB performance loss.

can achieve performance close to unquantized angles, with 0.1 dB performance loss.

can achieve performance close to unquantized angles, with 0.2 dB performance loss.

can achieve performance close to unquantized angles, with 0.1 dB performance loss.

can achieve performance close to unquantized angles, with 0.1 dB performance loss.

can achieve performance close to unquantized angles, with 0.1 dB performance loss.

can achieve performance close to unquantized angles, with 0.4 dB performance loss.

Simulation Results for Subcarrier Grouping APPENDIX B: Simulation Results for Subcarrier Grouping

Performance loss: Ng=2, 0.15 dB Ng=4, 0.2 dB Ng=6, 0.4 dB Ng=8, 0.8 dB

Performance loss: Ng=2, 0.2 dB Ng=4, 0.3 dB Ng=6, 0.5 dB Ng=8, 0.9 dB

Performance loss: Ng=2, 0.2 dB Ng=4, 0.3 dB Ng=6, 0.5 dB Ng=8, 1 dB

Performance loss: Ng=2, 0.2 dB Ng=4, 0.2 dB Ng=6, 0.5 dB Ng=8, 1 dB

Performance loss: Ng=2, 0.2dB Ng=4, 0.3 dB Ng=6, 0.6 dB Ng=8, 0.8 dB

Performance loss: Ng=2, 0.3 dB Ng=4, 0.7 dB Ng=6, 1.3 dB Ng=8, 2 dB

Performance loss: Ng=2, 0.2 dB Ng=4, 0.3 dB Ng=6, 0.6 dB Ng=8, 1 dB

Performance loss: Ng=2, 0.5 dB Ng=4, 0.8 dB Ng=6, 1.4 dB Ng=8, 2.1 dB

Performance loss: Ng=2, 0.5 dB Ng=4, 0.55 dB Ng=6, 0.7 dB Ng=8, 0.9 dB

Performance loss: Ng=2, 1 dB Ng=4, 1.2 dB Ng=6, 1.3 dB Ng=8, 1.5 dB

Performance loss: Ng=2, 0.6 dB Ng=4, 0.8 dB Ng=6, 1 dB Ng=8, 1.3 dB

Performance loss: Ng=2, 1.8 dB Ng=4, 2 dB Ng=6, 2.5 dB Ng=8, 3.4 dB

Thanks for Your Attention!