Improving the Resource Efficiency in Massive MIMO-NOMA Systems

This work discusses a cluster-based non-orthogonal multiple access (NOMA) structures aiming at improving the system energy efficiency (EE) and spectral efficiency (SE) in a Massive MIMO (M-MIMO) cell through user clustering and power allocation strategies. Effectively, this work proposes an improved clustering-based NOMA M-MIMO structure under the EE-SE perspective combining the EE and SE performance metrics. In the considered power-domain NOMA M-MIMO configuration, the users are grouped into several clusters; hence, each beamformer serves the users at a specific cluster. The EE-SE tradeoff, namely resource efficiency figure of merit in a cluster-based downlink NOMA M-MIMO is evaluated considering the application of two power allocation strategies. Based on two different channel cluster-head selection strategies, namely the equivalent channel vectors (Type I) and the channel matrix singular value decomposition (SVD) (Type II), the cluster-head in each cluster can tremendously mitigate the intra-cluster interference. Numerical results demonstrate that the best system performance regarding EE-SE tradeoff occurs for MIMO NOMA Type I cluster-head selection with two users per cluster.