Coordinated Multi-CPU for Interference Reduction in Cell-Free Massive MIMO-NOMA

The use of the power domain non-orthogonal multiple access (NOMA) in cell-free massive multiple-input multiple-output (CF-mMIMO) may further enhance spectral efficiency (SE). In this paper, we consider an environment with multiple central processing units (CPUs) to decentralize the processing on a CPU. However, when each CPU executes user clustering (UC) independently, the user terminal (UT) receives interference from access points (APs) controlled by the neighboring CPUs. To mitigate this interference, we propose and analyze a coordinated multi-CPU scheme in which CPUs share information from the APs. By sharing information from the APs, it is possible to control the APs of the neighboring CPUs that may interfere with UTs. Simulation results show that the coordinated multi-CPU reduces interference from APs connected to the neighboring CPUs and improves the SE compared to the uncoordinated multi-CPU under the same multi-CPU environment.