Power-Balanced Non-Orthogonal Multiple Access Based on Virtual Channel Optimization

In this brief, we propose a virtual channel optimization method for downlink non-orthogonal multiple access (NOMA) systems. With this method, we show that superposed users’ data can be separated successfully in power-balanced scenario for NOMA. We derive the optimum relative phase offset for two users as well as three users based on the minimum Euclidean distance (MED) among the superposed constellation points without channel state information at the transmitter (CSIT). We exploit maximum likelihood (ML) detector at the receiver as successive interference cancelation (SIC) is not applicable for power-balanced scenario. Since the complexity of ML is high, we design a low-complexity ML algorithm to substantially reduce the computational complexity. Simulation results for quadrature phase-shift keying (QPSK) modulation demonstrate the effectiveness and the performance of the proposed method.