A Novel Intelligent SIC Detector for NOMA Systems Based on Deep Learning

In this paper, we propose a novel intelligent successive interference cancellation (SIC) detection algorithm, namely I-SIC, for the uplink non-orthogonal multiple access (NOMA) system. Compared with some traditional SIC detection algorithms based on channel state information (CSI) and quality of service (QoS), the proposed I-SIC can learn the implied characteristics in the received signal, channel state information and power information through deep neural network (DNN), so as to more intelligently provide sorting scheme for SIC detection algorithm and further improve the detection performance of the system. Experimental results show that compared with the traditional SIC detection algorithm based on CSI (CSI-SIC), this algorithm can significantly improve the detection performance of the system(up to 6 dB for three-user scenario with QPSK modulation).