A Gradual Resource Allocation Technique for Massive MIMO-NOMA

In this letter, a new perspective for channel allocation and user pairing is proposed for a crowded downlink system using massive multiple-input–multiple-output, nonorthogonal multiple access (NOMA), and zero-forcing beamforming. The proposed user clustering technique allocates users to subbands and beams by a gradual relaxation of the correlation constraints in such a way to enhance the conditioning of the channel matrix. Channel measurements are performed in this context to validate the proposed technique in an experimental setup. Results show an important gain in performance in terms of both throughput and fairness with respect to orthogonal multiple access and to two state-of-the-art NOMA clustering methods.