A Random Fourier Feature Based Receiver Detection for Enhanced BER Performance in Nonlinear PD-NOMA

Non-orthogonal multiple access (NOMA) has been proposed as a potential enabler for massive connectivity. Among the different NOMA schemes, power-domain (PD-NOMA) is particularly appealing as it improves user fairness while enhancing spectral efficiency. However, the practical implementation of NOMA faces several challenges, including radio frequency impairments, such as power-amplifier (PA) nonlinearity, which can limit its performance. In this paper, we study the impact of PA nonlinearity on the detection performance of PD-NOMA, and propose a random Fourier feature (RFF) based solution to mitigate the effects of such imperfections. Computer simulations carried out assuming different PA nonlinearity types demonstrate that the proposed RFF based algorithm considerably improves the BER performance and achieves results that are close to the ideal case. Lastly, the analytical proof of our proposed algorithms performance is provided to support our simulation results.