Presentation is loading. Please wait.

Presentation is loading. Please wait.

1 Cross Evaluation of Proposed PHY Structures for the IEEE 802.16m UL Primary and Secondary Fast Feedback Channels Document Number: C802.16m-09/0125 Date.

Similar presentations


Presentation on theme: "1 Cross Evaluation of Proposed PHY Structures for the IEEE 802.16m UL Primary and Secondary Fast Feedback Channels Document Number: C802.16m-09/0125 Date."— Presentation transcript:

1 1 Cross Evaluation of Proposed PHY Structures for the IEEE 802.16m UL Primary and Secondary Fast Feedback Channels Document Number: C802.16m-09/0125 Date Submitted: 2009-01-05 Source Hongmei Sun, Changlong Xu, Jong-Kae (JK) Fwu, Email: {hongmei.sun, changlong.xu, jong-kae.fwu }@ intel.com Qinghua Li, Eddie Lin, Yuan Zhu, Hujun Yin, Roshni Srinivasan, Rath Vannithamby, Sassan Ahmadi Intel Corporation Venue: Re: 802.16m-08/052, Call for Comments on 802.16m SDD (802.16m-08/003r6), Section 11.9.2.1 Base Contribution: N/A Purpose: To be discussed and adopted by TGm for use in 802.16m SDD Notice: This document does not represent the agreed views of the IEEE 802.16 Working Group or any of its subgroups. It represents only the views of the participants listed in the “Source(s)” field above. It is offered as a basis for discussion. It is not binding on the contributor(s), who 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 802.16. Patent Policy: The contributor is familiar with the IEEE-SA Patent Policy and Procedures: and.http://standards.ieee.org/guides/bylaws/sect6-7.html#6http://standards.ieee.org/guides/opman/sect6.html#6.3 Further information is located at and.http://standards.ieee.org/board/pat/pat-material.htmlhttp://standards.ieee.org/board/pat

2 2 Contributions on Fast Feedback Channel Design EricssonC80216m-UL_PHY_Ctrl-08_08_062r1 & C80216m-UL_PHY_Ctrl-08_052.doc IntelC80216m-UL_PHY_Ctrl-08_065r2.ppt LGEC80216m-UL_PHY_Ctrl-08_039r1.pptC80216m-UL_PHY_Ctrl-08_039r1.ppt & C80216m-UL_PHY_Ctrl-08_FFBCH.ppt Mediatek (pilot design for control tile) C80216m-UL_PHY_Ctrl-08_060.ppt MotorolaSlides S802.16m-08/919Slides S802.16m-08/919r2 & C80216m-UL_PHY_Ctrl-08_058.ppt NextWaveC80216m-UL_PHY_Ctrl-08_038.doc NortelC80216m-UL_PHY_Ctrl-08_064.ppt SamsungIEEE C802.16m-08/982r2 C80216m-UL_PHY_Ctrl- 08_EnhancedFFBCH_Samsung.pdf

3 3 Outline Intel’s PCQICH design updated Fast Feedback Channel Design Cross Evaluation Results using evaluation criteria agreed upon in UL PHY Control RG –PCQICH: Intel, Samsung, MOT, Nortel –SCQICH: Intel, LGE, Mediatek, Ericsson, Nextwave, Samsung Conclusions and Recommendation

4 4 Intel’s Updated PCQICH Design PHY structure is as described in C80216m-UL_PHY_Ctrl- 08_065r2.ppt Dedicated coding for different information bits –Coding table size of [112x12] –4bits: code sequences 1~16: (max correlation distance = 1.95) –5bits: code sequences 17~48 (max correlation distance = 3.117) –6bits: code sequence 49~112 (max correlation distance = 3.86) Sequence mapping: –map each message to a group of codewords or code sequences and put each codeword in one FMT of the logical PCQICH channel –Can effectively improve performance at high speed: ex, Veh-A 350km/hr –Provides better performance than doing permutation proposed by LGE

5 5 Performance Comparison of PCQICH (Samsung, MOT, Nortel, Intel, LGE- option1) Note: LGE’s PCQICH-option2 design is not compared due to short notice.

6 6 PCQICH@PB3: 2x6 tile (Nortel 6x3) MOT: all results have been shifted left by 1.25dB(=10*log10(24/18)) to compensate for capacity difference 4bits: Intel, LGE outperform the rest 5bits/6bits: Intel outperforms the rest

7 7 PCQICH@ VA120 : 2x6 tile (Nortel 6x3) MOT: all results have been shifted left by 1.25dB(=10*log10(24/18)) to compensate capacity difference 4bits: Intel, Nortel, LG outperform others slightly 5/6bits: Intel’s design is slightly better than the rest

8 8 PCQICH@ VA350: 2x6 tile (Nortel 6x3) MOT: all results has been shifted left by 1.25dB(=10*log10(24/18)) to compensate capacity difference Nortel’s design outperforms others (by ~ 0.5dB) due to 6x3 tile vs. 2x6 tile W. filter based non-coherent detection: 2x6 tile can be used at 350kmph 2x6 tile: Intel’s design outperforms the rest Overall: Intel’s design is preferred (performance, legacy support, robustness)

9 9 PCQICH@VA350: 2x6 tile: sequence mapping (cont.) Sequence mapping: –Map each message to a group of codewords or code sequences and put each of the codewords to one FMT of the logical PCQICH channel Ex, 6bits inforation, code sequence index of each FMT will be 1) 1 st FMT: 2) 2 nd FMT: 3) 3 rd FMT: –This can effectively improve performance at high speed: ex,VA350 –Provides better performance than permutation proposed by LGe Conclusion –Sequence mapping improves the performance and lower the error floor –Should be used for enhanced basic receiver –Low pass filter can be used for advanced receiver

10 10 PCQICH@PB3: 6x2 tile (Nortel 6x3) 4bits: Intel’s design outperforms the rest 5/6bits: –Intel, Samsung are among the best Overall: Intel’s PCQICH design is preferred

11 11 PCQICH@ VA120 : 6x2 tile (Nortel 6x3) 4bits: Intel & Nortel slightly outperform the rest 5/6bits: Similar performance between designs from Intel, Nortel, & Samsung

12 12 PCQICH@ VA350 : 6x2 tile (Nortel 6x3) 4bits: Nortel slightly Intel & Nortel slightly outperform the rest 5/6bits: Intel and Samsung among the best Overall: 6x2 (Intel/Samsung) slightly outperform 6x3 for Veh A 350km/hr

13 13 Summary Scenarios4-bit Best Performance5/6-bit Best Performance PCQICH@PB3, 2x6 tileIntel, LGe-option1Intel PCQICH@VA120, 2x6 tileIntel, Nortel, LGe-option1Intel PCQICH@VA350, 2x6 tileNortel (6x3), Intel Summary Intel’s design is preferred (performance, legacy support and robustness) PCQICH@PB3, 6x2 tileIntelIntel, Samsung PCQICH@VA120, 6x2 tileIntel, NortelIntel, Samsung, Nortel PCQICH@VA350, 6x2 tileIntel, NortelIntel, Samsung Summary Intel’s design is preferred (performance, robustness, Legacy support) Note: LGE’s option-1 with 6x2 tile structure is not evaluated due to time constraints, but performance is expected to be similar to that of the 2x6 tile structure

14 14 Performance Comparison of SCQICH (Samsung, MTK, LGE, Ericsson, Nextwave, Intel)

15 15 SCQICH@AWGN: 2x6 tile, 7~12bits 7~12 bits SCQICH@AWGN: Eric ( TBCC, 1/5 rate), Samsung (block code), Intel have similar performance Same tile structure for the design, only AWGN is compared

16 16 SCQICH@AWGN: 2x6 tile, 24bits 24bits @AWGN: Ericsson (TBCC, rate 2/5) outperforms Intel Compared with Intel (single design for SCQICH), Ericsson’ solution has higher design complexity since different code design is used for the two ranges of information size for 13~24 bits and 7~12bits (Needs further analysis)

17 17 SCQICH@PB3: 2x6 tile 7bits: LGE, Intel & Samsung among the best 12/24bits: Intel/Samsung outperform LGE and Nextwave 24 bits Samsung design not compared yet. Overall: Intel, Samsung’s design provides best performance for Ped B 3km/hr

18 18 SCQICH@VA120: 2x6 tile 7bits: LGE, Intel & Samsung among the best 12/24bits: Intel & Samsung outperform LGe and Nextwave Samsung design with 24 bits not compared yet. Overall: Intel, Samsung’s design provides best performance @ Veh A 120km/hr

19 19 SCQICH@VA350: 2x6 tile Inner pilot pattern (from Samsung/MTK) is better than 2 pilots at corners 7bits: LGE, Intel & Samsung among the best 12/24bits: Intel & Samsung outperform LGe and Nextwave 24 bits Samsung design not compared yet. Overall: Intel & Samsung’s design provides best performance @ Veh A 350km/hr

20 20 SCQICH@PB3: 6x2 tile 7bits: LGE & Intel outperform Nextwave 12/24bits: Intel outperforms LGe Overall: Intel’s design has the best performance for Ped B 3km/hr

21 21 SCQICH@VA120: 6x2 tile 7bits: LGE & Intel’s designs outperform Nextwave’s design 12/24bits: Intel outperforms LGE Overall: Intel’s design provides best performance at Veh A 120km/hr

22 22 SCQICH@VA350: 6x2 tile 7bits: LGE & Intel outperform Nextwave 12/24bits: Intel outperforms LGe Overall: Intel’s design provides best performance at VA350

23 23 Summary Scenarios7-12bits Best Performance24bits Best Performance SCQICH @ AWGN, 2x6 tileEricsson, Intel, SamsungEricsson (Samsung’s TBCC: TBA) Scenarios7-bit Best Performance12/24-bit Best Performance SCQICH@PB3, 2x6 tileLGE, Intel, SamsungIntel, (Samsung)^^ SCQICH@VA120, 2x6 tileLGE, Intel, SamsungIntel, (Samsung)^^ SCQICH@VA350, 2x6 tileLGE, Intel, SamsungIntel, (Samsung)^^ Summary Single Design: Intel’s design is preferred (performance and robustness, simple design) Separate Design: Slight gain when optimized for different scenarios. SCQICH@PB3, 6x2 tileLGE, Intel, (Samsung)*Intel, (Samsung)^^ SCQICH@VA120, 6x2 tileLGE, Intel, (Samsung)*Intel, (Samsung)^^ SCQICH@VA350, 6x2 tileLGE, Intel, (Samsung)*Intel, (Samsung)^^ Summary Single Design: Intel’s design is preferred (performance and robustness, simple design) Separate Design: Slight gain when optimized for different scenarios. (Samsung)*: Does not includes results for 6x2 structure. Similar performance to 2x6 expected. (Samsung)^^: Samsung’s block code design has similar performance as Intel. 24bit design not compared yet.

24 24 Summary (cont.) PCQICH: –Intel’s design provided for 4bits, 5bits, 6bits. Design outperforms others in most scenarios –Sequence mapping can effectively lower the error floor for high speed scenario SCQICH: –7bits/12bits: Intel’s design has similar performance as LGE’s design –24bits: Intel’s design outperforms all others with a single code design –Separate code designs show small performance gain when optimized for different code rates (with added complexity).

25 25 Backup Slides

26 26 Simulation Setting Note: Low pass filter based non-coherent detection: the filter is built based on, where R is parameterized by Doppler speed. A fixed conservative time delay spread is used to get freq. domain correlation when calculating R since the channels are unknown. Channel Bandwidth10MHz Over-sampling Factor28/25 FFT Size1024 Cyclic prefix (CP) ratio1/8 Channel conditionPB3, VA120, VA350 The number of antennasTx:1, Rx:2 ModulationBPSK/QPSK Channel estimation2-D MMSE Tile size2x6, 6x2, 6x3 Block size6x6 ReceiverPCQICH: 6x2: non-coherent detection, MLD 2x6: low pass filter based non-coherent detection, MLD SCQICH: coherent detection, MLD

27 27 Low pass filter based non-coherent detection Detected CQI sequence where is the predefined orthogonal sequence, and r(j) is received signal of j-th subcarrier To reduce complexity, R can be predefined by assuming a fixed high speed and a conservative time delay spread without obvious performance loss

28 28 Semi orthogonal sequence of Intel’s updated PCQICH 4bits: 16 code sequences (max correlation distance = 1.95) 5bits: 32 code sequences (max correlation distance = 3.117) 6bits: 64 code sequences (max correlation distance = 3.86)


Download ppt "1 Cross Evaluation of Proposed PHY Structures for the IEEE 802.16m UL Primary and Secondary Fast Feedback Channels Document Number: C802.16m-09/0125 Date."

Similar presentations


Ads by Google