Performance of Uplink NOMA-MIMO System with Joint DPC-OSIC Detector

The enhanced capacity-achieving non-orthogonal multiple access (NOMA) systems suffer from detection errors due to the interference of various users in the uplink (UL) scenario. Successive interference cancellation (SIC) errors propagation and reception of signals from multiple antennas result in inter-user interference (IUI) and inter-antenna interference (IAI), respectively. Dirty paper coding (DPC) mitigates the IAI by diagonalizing the channel matrix, whereas SIC reduces the IUI by using the back-substitution method. This paper analyses the UL NOMA-MIMO system's performance by proposing a joint DPC-OSIC detector with known channel state information (CSI). QR-based OSIC reduces the IUI while applying all the constellations to the lowest channel gain user. The proposed joint DPC-OSIC is analyzed over a joint DPC-ZF and DPC-ML performance. Simulation results validate that the proposed detector achieves near maximum likelihood (ML) performance with relatively low complexity.