On approximated LLR of MMSE equalizer for millimeter-wave in presence of phase noise

This paper proposes an approximated log likelihood ratio (LLR) for millimeter-wave (mmW) systems in presence of phase noise. In mmW systems, phase noise of carrier wave signal causes severe performance degradation while operating in much higher frequency. In order to mitigate the impairments of phase noise, space-time block code (STBC) and forward error correction (FEC), such as LDPC, techniques are effective tools. However, if the probabilistic model does not capture exact behavior of a random process on received signal, the capability of FEC is severely deteriorated, especially in higher order modulation or higher coding rate cases. To address this issue, we define the probabilistic model of minimum mean square error (MMSE) equalizer output including phase noise components. Then, a method for approximately computing the LLR is proposed by applying Gaussian approximation. We have demonstrated that the approximated LLR computation from the defined probabilistic model is capable of improving bit error rate (BER) performances in the presence of phase noise.