MMSE Hybrid Beamforming for Weighted Sum Rate Maximization in NOMA Systems

This work proposes a multiuser hybrid beamforming scheme for non-orthogonal multiple access (NOMA) systems using the minimum mean square error (MMSE) approach to weighted sum rate maximization. While NOMA may be effective in terms of enhancing user fairness, hybrid beamforming is necessary to reduce the transceiver cost as the system moves towards higher frequency. The design is divided into two stages. In the first stage, a fully digital multiuser beamformer is derived by maximizing the weighted sum rate of all users under no constraint on the number of RF chains. This problem is then transformed into a weighted sum MSE minimization problem, which facilitates the use of alternating optimization to obtain an efficient local solution. In the second stage, the previously obtained multiuser beamformer is then split into RF and baseband beamformers by using orthogonal matching pursuit (OMP). A user role selection algorithm is then proposed to determine the role of strong and weak users. Simulation results are provided to demonstrate the effectiveness of the proposed schemes.