1. Submission doc.: IEEE 11-13/1059r0 September 2013 Dongguk Lim, LG ElectronicsSlide 1 PHY Abstraction for HEW Evaluation Methodology Date: 2013-09-11 Authors:

2. Submission doc.: IEEE 11-13/1059r0 Introduction The objective of PHY abstraction is to accurately predict link layer performance in computationally easy way This contribution compares two PHY abstraction methods: Mutual Information based approach (MMIB) Constrained Capacity based approach (CC) Slide 2Dongguk Lim, LG Electronics September 2013

3. Submission doc.: IEEE 11-13/1059r0 System model Effective SINR (SINR eff ) can be calculated as follows where SINR n is the post processing SINR at the n-th subcarrier, N is the number of symbols for a coded block or the number of data subcarriers used in an OFDM system, and Φ is Effective SINR Mapping (ESM) function Since the same SINR eff provides the same Packet Error Rate (PER) performance for a given coding block size and MCS level, we can use AWGN performance curve as a reference curve Slide 3Dongguk Lim, LG Electronics September 2013

4. Submission doc.: IEEE 11-13/1059r0 SINR n Calculation (Example) For SISO case, y= hx+n where y is a received signal h is channel response at each subcarrier x is a transmitted signal n is a noise Then, SINR n can be calculated as where ɛ x is a signal strength σ n 2 is noise variance Slide 4Dongguk Lim, LG Electronics September 2013

5. Submission doc.: IEEE 11-13/1059r0 Code Block Size Different code block sizes provide different PER performances Ideally, we need to have AWGN curves for all possible code block sizes Practically, we consider the following approach for simplicity Produce reference curves for several code block sizes Do interpolation for other code block sizes Slide 5Dongguk Lim, LG Electronics September 2013

6. Submission doc.: IEEE 11-13/1059r0 Two ESM functions Slide 6Dongguk Lim, LG Electronics September 2013 Mutual Information based approach (MMIB) [1] Constrained Capacity based approach (CC) [2]

7. Submission doc.: IEEE 11-13/1059r0 Mutual Information based approach (MMIB) (1/3) Each bit experiences a different ‘equivalent’ bit-channel Due to the asymmetry of the modulation map Mutual information of the equivalent channel is where m is the number of bits per constellation, and I(b i,LLR(b i )) is the mutual information between input bit to the QAM mapper and output LLR for i-th bit in the modulation map Slide 7Dongguk Lim, LG Electronics September 2013 MMIB: Mean Mutual Information per coded Bit

8. Submission doc.: IEEE 11-13/1059r0 Mutual Information based approach (MMIB) (2/3) Mean mutual information through N sub-carriers over the codeword Since mutual information I(b i,LLR(b i )) is a function of constellation and SINR, mean mutual information is Slide 8Dongguk Lim, LG Electronics September 2013

9. Submission doc.: IEEE 11-13/1059r0 Mutual Information based approach (MMIB) (3/3) Effective SINR mapping (ESM) function is derived for each modulation as follows (details in [1]) Slide 9Dongguk Lim, LG Electronics September 2013 ModulationNumerical Approximation BPSKK=1, a = [1], c = [2√2] QPSKK=1, a = [1], c = [2] 16-QAMK=3, a = [0.5 0.25 0.25], c = [0.8 2.17 0.965] 64-QAMK=3, a = [1/3 1/3 1/3], c = [1.47 0.529 0.366]

10. Submission doc.: IEEE 11-13/1059r0 Constrained Capacity based approach (CC) Simply use capacity formula for ESM function where M is the maximum number of bits for all constellation CC is less complex than MMIB since there is a closed form solution of Φ inverse Slide 10Dongguk Lim, LG Electronics September 2013

11. Submission doc.: IEEE 11-13/1059r0 Performance Comparison: QPSK Slide 11Dongguk Lim, LG Electronics September 2013

12. Submission doc.: IEEE 11-13/1059r0 Performance Comparison: 16QAM Slide 12Dongguk Lim, LG Electronics September 2013

13. Submission doc.: IEEE 11-13/1059r0 Conclusion This contribution provided introduction on PHY Abstraction We compared two methods and observed that MMIB provides accurate prediction of link performance CC requires less computational complexity, but shows some inaccuracy For accurate prediction of link performance, MMIB should be adopted as one of PHY abstraction methods Slide 13Dongguk Lim, LG Electronics September 2013

14. Submission doc.: IEEE 11-13/1059r0 Reference [1] IEEE 802.16m-08/004r5, “IEEE 802.16m Evaluation Methodology Document (EMD)” [2] IEEE 802.11-13/0757r1, “Evaluation Methodology and Simulation Scenarios” Slide 14Dongguk Lim, LG Electronics September 2013

15. Submission doc.: IEEE 11-13/1059r0 Appendix MMIB Performance Verification May 2013 Dongguk Lim, LG ElectronicsSlide 15

16. Submission doc.: IEEE 11-13/1059r0 Simulation Parameters Basic parameters To eliminate the channel impairment effect such as ICI, ISI, channel estimation error, we used the perfect channel estimation and increased CP length in simulation Effect of channel impairments are for further study Slide 16Dongguk Lim, LG Electronics September 2013 Frequency band2.4 GHz Band Width20 MHz FFT Size64 Channel ModelAWGN, UMi, UMa Channel conditionLOS, NLOS Channel EstimationPerfect PHY Abstraction methodMMIB, CC

17. Submission doc.: IEEE 11-13/1059r0 MMIB MCS0 Slide 17Dongguk Lim, LG Electronics September 2013

18. Submission doc.: IEEE 11-13/1059r0 MMIB MCS1 Slide 18Dongguk Lim, LG Electronics September 2013

19. Submission doc.: IEEE 11-13/1059r0 MMIB MCS2 Slide 19Dongguk Lim, LG Electronics September 2013

20. Submission doc.: IEEE 11-13/1059r0 MMIB MCS3 Slide 20Dongguk Lim, LG Electronics September 2013

21. Submission doc.: IEEE 11-13/1059r0 MMIB MCS4 Slide 21Dongguk Lim, LG Electronics September 2013

22. Submission doc.: IEEE 11-13/1059r0 MMIB MCS5 Slide 22Dongguk Lim, LG Electronics September 2013

23. Submission doc.: IEEE 11-13/1059r0 MMIB MCS6 Slide 23Dongguk Lim, LG Electronics September 2013

24. Submission doc.: IEEE 11-13/1059r0 MMIB MCS7 Slide 24Dongguk Lim, LG Electronics September 2013