Deep Learning-Based Joint Detection for OFDM-NOMA Scheme
Yihang Xie, Kah Chan Teh, Alex C. Kot
Abstract
Non-orthogonal multiple access (NOMA) technique has drawn much attention in recent years. It has also been a promising technique for the fifth-generation (5G) wireless communication system and beyond. In this letter, we develop a novel deep learning (DL) aided receiver for NOMA joint signal detection. The DL-based receiver serves as an end-to-end mode, which simultaneously fulfills the function of channel estimation, equalization, and demodulation. Compared with the traditional signal detection method for the NOMA scheme, the proposed deep learning method shows feasible improvement in performance and robustness with the tapped-delay line (TDL) channel model, which is adopted for the 5G communication environment.