1. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 1 Update on MITMOT IEEE802.11n proposal Notice: This document has been prepared to assist IEEE 802.11. 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 grants a free, irrevocable license to the IEEE to incorporate material contained in this contribution, and any modifications thereof, in the creation of an IEEE Standards publication; to copyright in the IEEE’s name any IEEE Standards publication even though it may include portions of this contribution; and at the IEEE’s sole discretion to permit others to reproduce in whole or in part the resulting IEEE Standards publication. The contributor also acknowledges and accepts that this contribution may be made public by IEEE 802.11. Patent Policy and Procedures: The contributor is familiar with the IEEE 802 Patent Policy and Procedures, including the statement "IEEE standards may include the known use of patent(s), including patent applications, provided the IEEE receives assurance from the patent holder or applicant with respect to patents essential for compliance with both mandatory and optional portions of the standard." Early disclosure to the Working Group of patent information that might be relevant to the standard is essential to reduce the possibility for delays in the development process and increase the likelihood that the draft publication will be approved for publication. Please notify the Chair as early as possible, in written or electronic form, if patented technology (or technology under patent application) might be incorporated into a draft standard being developed within the IEEE 802.11 Working Group. If you have questions, contact the IEEE Patent Committee Administrator at.http:// ieee802.org/guides/bylaws/sb-bylaws.pdfstuart.kerry@philips.compatcom@ieee.org Date: 2005-07-18 Authors: “Mac and mImo Techniques for MOre Throughput”

2. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 2 Authors, contd:

3. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 3 Overall goal and positioning Preserve compatibility with legacy IEEE802.11 system Evolution: expand current WLAN application domain, offer a consistent solution to –Provide required QoS to support consumer electronics (multimedia home environment and VoIP enterprise) –Grant range extension for limited outdoor operation (hotspot) as well as full home coverage –Support heterogeneous traffic: increase overall peak data rate without jeopardizing lower data rates modes –Manage diversity (laptop/PDA/VoIP Phone) and evolution (independent STA/AP antenna configuration upgrade) of devices through asymmetric antenna configurations Proven and simple solution: combine a highly efficient contention-free based MAC with robust yet low complexity open-loop MIMO PHY techniques

4. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 4.11n MAC: an evolutionary approach Solutions: –Centralised on demand resource allocation with grouped resource announcements, embedded in.11e superframe providing contention free access for all type of.11n traffics –Aggregated PHY bursts made of short fixed size MAC-PDUs allows 1 or multiple destinations and/or PHY modes –Enhanced ACK: consists in a low latency and low overhead selective retransmission Benefits: –Actual QoS: guaranteed throughput, stringent delay constraints support even in heavily loaded system –High efficiency and scalable architecture scenario SS16 (point to point): 86% - extended SS6 (Hotspot): 67% maintain constant overhead when data rate increases –Efficient for heterogeneous traffics (bursty, VBR, CBR, high or low data rates) without parameter tuning –Easy implementation, low power consumption

5. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 5.11n PHY: a MIMO robust extension Goal: define new OFDM MIMO modes with the constraints to –handle asymmetric TX/RX antenna configurations with 1, 2 or 3 parallel streams –focus on open-loop for stability, avoiding calibration circuit or feedback signaling Solution: exploit a hybrid combination of –spatial Division Multiplexing (SDM) to increase spectrum efficiency and peak data rates –classical Space Time Block Coding (STBC) to improve link robustness or range for low to medium data rates (suited to small packet size e.g. VoIP) Additional key features: –mandatory: 20MHz bandwidth, minimum of 2Tx antennas (up to 4Tx) –new two stage space and frequency interleaver design –Forward Error Correction scheme: supports all.11a CC rates, adds low redundancy 5/6 (mandatory) advanced optional scheme: duo-binary turbo code similar as in IEEE802.16 –second 20MHz/128 carriers OFDM modulation (8% rate increase), with double duration guard interval (Hotspot: limited outdoor) –optional high rate 40MHz bandwidth/128 carriers modes (117% rate gain) –new nPLCP preambles: code overlay STS/orthogonal LTS combined with CS –optional simple PAPR reduction method based on pilot tones rotation for low power

6. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 6 PHY presentation outline The MitMot PHY layer proposal consists in an extension of IEEE802.11a PHY including several key new features: 20MHz (mandatory), 40MHz (optional) bandwidth Optional second OFDM modulation using 104 data subcarriers among 128 in 20MHz or 40MHz bandwidth Multiple TX/RX antenna modes handling asymmetric antenna configuration ( 2, 3 or 4 transmit antennas, 2 or more receiving antennas) Frequency and spatial interleaving Advanced optional forward error correction scheme relying on turbo-codes Improved preamble design for multi-antenna channel estimation and synchronization purposes Optional PAPR reduction method based on pilot tones rotation Link quality metric feedback for efficient link adaptation Simulation Results & Conclusion

7. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 7 OFDM modulations 2 bandwidths support: 20&40MHz 1 st OFDM modulation based on IEEE802.11a for 20MHz –48 data subcarriers, 64-point (I)FFT, 4 pilots  Reference PHY rate: 2TX: 120/144Mbps, 3-4TX: 180/216Mbps 2 nd OFDM modulation for 20MHz (optional): –duration of the guard interval and number of carriers doubled (0.8µs  1.6µs) to absorb larger multipath delays with same total overhead (25%) –104 data subcarriers, 128-point (I)FFT, 8 pilots  8% increase on PHY rate: 2TX: 130/156Mbps, 3-4TX: 195/234Mbps 3 rd OFDM modulation for 40MHz (optional): –104 data subcarriers, 128-point (I)FFT, 8 pilots –Guard interval duration: 0.8  s  117% increase on PHY rate: 2TX: 260/312Mbps, 3-4TX: 390/468Mbps

8. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 8 Multi-antenna scheme  Proposition: hybrid schemes relying on a combination of robust Space-Time Block Coding (STBC) and Spatial Division Multiplexing (SDM) –very simple transmitter implementation –very simple receiver implementations are possible as classical orthogonal designs are part of the proposed STBCs e.g. design of low complexity ZF or MMSE equalizers –very good performance complexity tradeoff for robustness in asymmetric MIMO Importance of configurations in which N Tx ≠ N Rx –N Tx > N Rx e.g. between AP and mobile handset (in DL) –N Tx < N Rx e.g. between MT and AP (UL), or if MT have upgraded multi-antenna capabilities compared to AP (infrastructure upgrade cost) –Exploit all available transmit diversity when N Tx > N Rx to improve Tx reliability 2, 3 or 4 transmit antennas –The number of receive antennas determines the maximum number of spatial streams that can be transmitted. –The capability of decoding 2 parallel data streams is mandatory –all the devices have to be able to decode all the modes where the number of spatial streams is lower or equal than the number of receive antennas in the device. –It is required for a device to exploit all its antennas in transmission even for optional modes. 2 or more receive antennas –With STBC or STBC/SDM, asymmetric antenna configurations can be supported Construction: transmission of 1, 2 or 3 parallel streams using,

9. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 9 2 transmit antenna schemes proposed3 transmit antenna schemes proposed 4 transmit antenna schemes proposed Asymmetric Modes for a robust hybrid solution

10. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 10 Asymmetric MIMO motivation/illustration Simulation parameters 20MHz bandwidth, 48 carriers 64QAM, CC 2/3 and 5/6 Packet size: 1000 bytes Channel TGn D NLOS MMSE MIMO detection, perfect CSI 2x2  3x2: 2.8dB  5dB gain @PER=10 -2 3x3  4x3: 2dB  5.4dB gain @PER=10 -2 2x2  4x2: 4.3dB  7.5dB gain @PER=10 -2

11. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 11 Frequency and spatial interleaver 2-step interleaving process Interleaving prior to mapping –802.11a like frequency interleaving with new parameters suitable to both OFDM modulations (48 and 104 subcarriers) Interleaving prior to space-time coding –based on the frequency interleaver parameters to ensure adjacent bits are transmitted on different streams N SD : number of data subcarriers

12. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 12 PHY modes & Environment/Device/Appl target

13. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 13 Forward Error Correction scheme Proposed FEC scheme: –Mandatory: all IEEE802.11a CC with 5/6 puncturing pattern –Optional: Advanced coding scheme Dual-Binary Turbo Code (Convolutional Turbo Code) –Dual-Binary is state of the art in Turbo Codes –Extremely good performance at small block sizes (excellent for larger as well) –TCs are stable, well-understood with excellent performance and known IPR landscape Complexity: –High parallelism possible without noticeable impact on performance –Easily adaptable to both low and high-end devices –Parallelism of 6, 8, 12, and 16 can enable most HT data-rates and satisfy SIFS. –Ex. Handset : 260 k Gates. P=6, N it = 5, 240 Mbps max throughput! –Ex. Set Top Box: 444 k Gates. P=12, N it = 5, 480 Mbps max throughput!

14. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 14 Duo-Binary Turbo Code Stucture Algorithmic interleaving scheme, defined by one equation and 4 parameters ─ Can ajust to any blocksize (only the 4 parameters are affected) ─ inherent parallelism (up to 16 currently) ─Specifically designed to be contention free ─No ROM, no routing issues (quasi regular) High flexibility: same encoder/decoder structure for any block size and coding rate (puncturing)

15. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 15 Gain of Duo-Binary Turbo Codes Turbo vs. Convolutional Codes AWGN 8000 info bits 2 to 2.5 dB gain @ PER 1% TC, 5 vs. 8 iterations AWGN Coded Block size of 648 bits <= 0.1 dB loss @ PER 1%

16. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 16 PAPR reduction by pilot rotation Motivation: Peak-to-Average-Power Ratio (PAPR) is an inherent OFDM issue occurring for all IEEE802.11n proposals –PAPR reduction: reduced constraints on PA back-off  gain in power @ constant out-of-band radiation or performance (limitation of non linearities) Constraints for MitMot proposal: –no signalization (avoid throughput loss) –no reservation of tones, no oversampling requirements –plug-in approach: ensure compatibility with all proposals Proposal: introduce additional degree of freedom on pilot design –preserve the 4 (#48c)/ 8 (#104c) fixed pilots –if PAPR is above defined threshold (7dB), TX chooses these pilots from QPSK (BPSK) alphabet such that resulting time domain PAPR is minimum –at RX effect of unknown rotation is cancelled through simple phase multiplication and classical phase tracking procedure is applied Definition of PAPR:

17. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 17 PAPR: MIMO Principles & Performance Encoding/Decoding in MIMO context: –For TX / RX operations: near optimum algorithms can be applied at very low complexity In TX, only additions in combination with a look-up table are required –Detection in RX: In combination with STBC and/or SDM, two principles may apply: Test all possible permutations (ML detection) Perform SDM/STBC decoding on pilots and perform modulo operation Simulation results: –In a typical example (STBC 2x2, QAM16 constellations, PA impairments, BPSK pilots): No performance degradation due to pilot choices PAPR reduction approx. 1.25dB (1.5dB with QPSK pilots) Gain achieved through PAPR reduction: –Decrease constraints on P1dB / back-off leads to a reduced power consumption Extrapolation of data provided by various vendors: 2dB gain in Op1dB yielding 26% power saving on PA

18. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 18 Preamble Design nSTS : time synchronization, frequency offset, AGC –Code overlay time domain sequence design on finite alphabet {0,±1, ±j} leads to simple cross correlator implementation nLTS : synchronization refinement, channel estimation –Orthogonal design (Walsh-Hadamard weighting) combined with cyclic shift approach: improves system by reducing RX power variations related to de/constructive recombination effects

19. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 19 Short Training Sequence Preamble Observation: STS are mainly used for AGC settling, time offset and rough frequency offset estimation. A pure time domain challenge! Question: why adopting a frequency domain specification of these sequences? MITMOT approach: perform time domain design using alphabet {0,±1, ±j} –nSTS choice criteria: Spectral & auto/cross-correlation properties Benefit: enables low PAPR LTS, yields extremely simple crosscorrelator implementation for time synchronization Frame definition: nSTS are weighted by ±1, nSTS doubled @40MHz

20. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 20 Long Training Sequence preamble Focus on a orthogonal design allowing –easier tradeoff between quality/complexity for CSI estimation: frequency domain only estimation is possible –Inclusion of time confinement constraint into the estimator possible yielding a more robust estimator avoiding the important noise enhancement using ZF approaches with Cyclic Shift based methods Definition in frequency domain from alphabet {0, ±1} –LTS over 56 subcarriers to further improve the accuracy of the channel estimator using time confinement constraint LTS(#-28…#+28) = {-1, 1, -1, 1, 1, 1, 1, -1, -1, 1, 1, 1, -1, 1, 1, -1, -1, -1, -1, 1, 1, -1, 1, 1, -1, 1, -1, -1, 0, -1, -1, -1, 1, -1, 1, -1, -1, -1, 1, 1, 1, 1, -1, -1, 1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -1, 1, -1}

21. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 21 Limited outdoor environment: Hotspot support Benefit of 20MHz 128 carriers mode using a 32 samples cyclic prefix: –8% overall rate increase –designed to cope with larger channels for more efficient outdoor environment operations Illustration for channel F: no error floor in performance for higher rates modes

22. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 22 Topology of measurements Map of test area with distances 1: 600ft 2: 1100ft 3: 1650ft 4: 250ft 5: 850ft 6: 1700ft 7: 300ft 8: 500ft 9: 450ft 10: 250ft

23. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 23 Delay Spread Measurements + RMS delay spread Delay spreads for most near-range cases #4,#5 and #10 Conclusions –Channel contributions approx. limited to GI size of 800ns, except for #10: a considerable RMS delay spread length is observed (> 1200ns with proba 25%) –Necessity of a larger GI ?  SNR degradation due to IBI

24. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 24 Impact of IBI on link performance Working assumptions: –Evaluation based on TGn-D channel simulation results –Effective SNR derived by considering IBI for 1600ns of delay spread (delay profile from scenario #10) and thermal noise Comparison with and w/o IBI for 2x2 case Conclusion: –Highest throughput modes are no longer available with IBI (noise + interf. saturation) –Communication range is decreased for all modes (decrease gets stronger with SNR requirements), e.g. 54Mbps: 42.9m range w/o IBI  38.1m with IBI

25. doc.: IEEE 802.11-05/0735r0 Submission July 2005 Marc de Courville - Motorola, John Benko - FTSlide 25 Differentiators Capture wide range of environments/devices/applications: –(full) home/enterprise/limited outdoor, handhelds/laptops, from VoIP to multimedia streaming –Build in support for asymmetric TX/RX antenna configurations to accommodate various terminal sizes (PDA/Phone) offering a scalable and evolutionary solution –Hotspot support: dedicated 128 carrier with double length cyclic prefix OFDM modulation, longer range achieved through hybrid STBC robust and SDM high peak rates modes –.11n specific robust beacon enables materialization of new PHY mode range prediction Enhanced QoS using “ Extended Centralized Coordination Function” –Inherent clean split between legacy and.11n devices at MAC level with no need for mixed-modes transmission mode definition –Resource allocation mechanism is highly dynamic QoS provided without use of traffic profiles (TSPECS) –High Efficiency independent of application packet size through segmentation –Robustness to error through retransmission mechanism on segmented packets Lower power operation: –PHY power saving: PAPR reduction based on simple pilot rotation –Enhanced transparency and predictability through broadcast grouped resource announcement yields clean low power implementation and low overhead New preamble definition: for simple&accurate AGC, time sync and easier quality/complexity tradeoff for CSI estimation Improved link adaptation: efficient interoperability through calibration and support of accurate link quality metrics