LLR Calculation Based on Interference Cancelation with Channel Estimation Error for Non-Orthogonal Multiple Access

Non-orthogonal multiple access improves spectral efficiency through sharing the same frequency resource with multiple wireless terminals. In previous researches, it has been shown that a log likelihood ratio (LLR) calculated from a signal on the sharing spectrum should be masked to zero if the power of an interference signal is stronger than that of a desired signal. This implies that the spectrum is not shared by the terminal receiving the weaker signal. In this paper, a spectrum sharing scheme with the use of interference cancelation and LLR calculation on the basis of the probability density function of residual interference is proposed. The residual interference remains owing to a channel estimation error and the LLR calculation takes that into account. Numerical results obtained thorough computer simulation show that the proposed scheme improves the performance by 1 dB with ideal channel estimation and 2 dB with channel estimation error at the BER of 10^-4 as compared to that with a conventional LLR calculation scheme.