Proposition of a Novel Clustering Methodology for Multi-User MIMO-PD-NOMA

Non-Orthogonal Multiple Access (NOMA) in association with Multiple-Input-Multiple-Output (MIMO) serves to improve the system's comprehensive performance, specifically in regard to spectral efficiency and reliability, in consort with reducing latency, thereby making it a more propitious candidate for beyond 5G networks. Clustering methodologies have garnered substantial consideration in the domain of wireless communication owing to their capability to improve allocation of resources, efficient management of the network, and comprehensive system performance. NOMA uses Successive Interference Cancellation (SIC) technique that tends to get more imperfect with the enhancement in the participating users in the network; therefore, applying clustering aids to be advantageous in such scenarios, wherein users can be assembled into small clusters, with each cluster employing the NOMA principle independently. This paper proposes a novel clustering methodology whose performance is scrutinized in comparison with some of the popular clustering techniques, for a downlink network scenario, in conjunction with MIMO Power-Domain NOMA (MIMO-NOMA), involving Constrained Optimization Power Allocation (COPA) mechanism. The performance of different clustering methodologies is analyzed using a multi-user MIMO-NOMA (MV-MIMO-NOMA) platform, with respect to Sum Capacity, system Bit Error Rate (BER) and varied modulation schemes.