Massive MIMO NOMA Downlink

The performance of multi-cell massive multiple-input multiple-output (MIMO) downlink (DL) with non-orthogonal multiple access (NOMA) is investigated. A novel pilot assignment for NOMA users is proposed to mitigate intra-cluster pilot contamination and to alleviate the detrimental impact of inter-cluster in the useful operating regime. The user nodes grouped into the same cluster based on the spatial-direction information are assigned orthogonal pilots to mitigate intra-cluster pilot contamination. These pilots are shared among the clusters to reduce the training overhead. However, the inter-cluster pilot contamination is mitigated by exploiting the distinct spatial directions of clusters. This pilot assignment ensures that the number of NOMA users that can be served simultaneously in the same time-frequency-spatial resource block is not limited by the length of coherence interval. A max-min fairness optimal transmit power allocation algorithm is proposed to guarantee user-fairness. The achievable DL sum rates are derived, and thereby, the cumulative impact of channel estimation errors, imperfect successive interference cancellation, and statistical channel knowledge at users is quantified. Our analysis reveals that the proposed massive MIMO NOMA can support massive access without compromising the spectral efficiency gains.