A Proportional Fairness-based Power Allocation Scheme for Non-orthogonal Multiple Access Downlink Systems

This paper tackles the problem of proportional fairness-based power allocation in downlink non-orthogonal multiple access (NOMA) systems. We address the scenario where the users are divided into groups of two. NOMA technique is applied within each group with the superposition coding at the transmitter side and successive interference cancellation (SIC) at the receivers' side, and different groups are assigned separate channels. We assume that the grouping and the channel assignment are dynamically performed, and we optimize the power allocation among users by maximizing the sum of logarithmic rates with a constraint on the total transmit power. We prove that the globally optimal solution to the optimization problem can be obtained by solving the Karush-Kuhn-Tucker (KKT) conditions and propose an efficient algorithm for this purpose. Extensive experimental results demonstrate that the proposed scheme outperforms several existing NOMA schemes in terms of system throughput and/or fairness, achieving the best trade-off between these two criteria.