A novel low-complexity detection scheme for power-domain NOMA systems

Non-orthogonal multiple access (NOMA) offers a solution to the increasing demands on the spectrum resources by allowing multiple users to occupy the same channel, significantly improving the spectral efficiency. However, one of NOMA's main challenges is the high computational complexity (CC) induced in the detection process. In this paper, a novel low-complexity detector for power-domain NOMA systems is proposed. The proposed scheme relies on a bit-wise symbol detection (rather than conventional symbol-wise detection), which significantly reduces the CC without any effect on the bit error performance. The performance of the proposed detector is analyzed and compared with the optimal maximum-likelihood detector and the conventional successive interference cancellation one, where analytical and simulation results reveal that the three schemes can achieve the same error performance, while much lower CC is required in the proposed detector compared to the two other baseline ones.