Answers are grounded in this paper and cite their sources.
Serial interference cancellation (SIC) is a conventional demodulation method for Power domain non-orthogonal multiple access (PD-NOMA) in vehicular communications. In this article, we derived that the traditional SIC method causes inter user crosstalk under specific power allocation strategies, resulting in the inability to correctly decode bit information. We call this phenomenon serial demodulation crosstalk (SDC). In addition, when the channel conditions change, the receiving user is unable to quickly obtain the changes of the power allocation strategy, which might also lead to error demodulation. To address these issues, one Deep-Learning Neural Network (DNN)-assisted anti-crosstalk parallel demodulation method is proposed. After receiving the superimposed signals from the base station, the users can employ a shallow DNN to identify dynamic power ratios in a very short time. Then, the current user information can be demodulated with the proposed anti-SDC parallel demodulation algorithm based on this power ratio. The experimental results show that the proposed DNN-assisted anti-crosstalk parallel demodulation method can adapt to various power ratio states and achieve high-precision demodulation with low computational complexity in the case of high modulation orders.
No takes yet. Share an insight, caveat, or question.
Li et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: