Polar coded MQAM with no noise variance estimation for capacity and soft decision metric

Polar coding (PC) separates the codeword into free and frozen sets of channels. In this paper, the construction of the two sets of channels and decoding are carried out based on the channel capacity and soft decision metric, respectively with unity noise variance. Additionally, the capacity for which signal and noise power are equal results in an interesting capacity spectrum whose average value is used in the PC construction. Simulations results show that this simplified soft metric achieves the same bit error rate (BER) when compared to the exact metric with receive overhead advantage. Transmission of symbols is made via flat fading Rayleigh channels. We see that the simulations results are well bounded within the theoretical performance limits.