A Two-Layer RSMA Framework With Balanced Clustering Design for Cell-Free Massive MIMO Systems

In this article, a 2-layer rate-splitting multiple access (RSMA) framework with a balanced clustering design is proposed for cell-free massive multiple-input multiple-output (CF-mMIMO) systems. Aiming at enhancing spectral efficiency (SE) and meanwhile ensuring low-complexity and scalability in large-scale systems, this work addresses different types of multiuser interferences by utilizing a two-stage optimization scheme with a designed spatial reduction matrix that encompasses the clustering design and joint RSMA. First, a balanced clustering design is developed for 2-layer RSMA to manage intra- and intercluster interference more efficiently than conventional user-centric clustering, simultaneously considering access point (AP)-user connectivity in CF-mMIMO systems and the impact of cluster similarity on RSMA. By employing the spectral clustering method to solve the bipartite graph partitioning problem with the min-max cut objective, the number of clusters is determined adaptively. Based on the above clustering design, a joint optimization of inner and outer RSMA is proposed to mitigate intra- and intercluster interferences simultaneously. Simulation results verify the SE enhancement, low-complexity, and scalability of the proposed 2-layer RSMA framework, compared with benchmark frameworks.