PHY Abstraction for HEW Evaluation Methodology

PHY Abstraction for HEW Evaluation Methodology • Date:2013-09-11 Authors: Dongguk Lim, LG Electronics

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) Dongguk Lim, LG Electronics

System model • Effective SINR (SINReff) can be calculated as follows • where SINRn 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 SINReff 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 Dongguk Lim, LG Electronics

SINRn 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, SINRn can be calculated as • where ɛxis a signal strength • σn2is noise variance Dongguk Lim, LG Electronics

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 Dongguk Lim, LG Electronics

Two ESM functions • Mutual Information based approach (MMIB) [1] • Constrained Capacity based approach (CC) [2] Dongguk Lim, LG Electronics

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(bi,LLR(bi)) is the mutual information between input bit to the QAM mapper and output LLR for i-th bit in the modulation map MMIB: Mean Mutual Information per coded Bit Dongguk Lim, LG Electronics

Mutual Information based approach (MMIB) (2/3) • Mean mutual information through N sub-carriers over the codeword • Since mutual information I(bi,LLR(bi))is a function of constellation and SINR, mean mutual information is Dongguk Lim, LG Electronics

Mutual Information based approach (MMIB) (3/3) • Effective SINR mapping (ESM) function is derived for each modulation as follows (details in [1]) Dongguk Lim, LG Electronics

Constrained Capacity based approach (CC) • Simply use capacity formula for ESM function • where Mis the maximum number of bits for all constellation • CC is less complex than MMIB since there is a closed form solution of Φ inverse Dongguk Lim, LG Electronics

Performance Comparison: QPSK Dongguk Lim, LG Electronics

Performance Comparison: 16QAM Dongguk Lim, LG Electronics

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 Dongguk Lim, LG Electronics

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” Dongguk Lim, LG Electronics

Dongguk Lim, LG Electronics Appendix MMIB Performance Verification

SimulationParameters • 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 Dongguk Lim, LG Electronics

MMIB MCS0 Dongguk Lim, LG Electronics

PHY Abstraction for HEW Evaluation Methodology