Joint User Pairing and Resource Allocation Optimization in Downlink 2-Layer Cooperative RSMA Networks

This paper introduces a 2-layer cooperative rate-splitting multiple access (C-RSMA) framework designed for multiple groups of two users. Within each user group, the message is divided into three components: an inter-group common message, an inner-group common message, and a private message. Our framework incorporates a novel user-pairing policy, leveraging a combination of semi-orthogonal user selection (SUS) and a matching-game (MG)-based algorithm to identify user pairs, which allows for selecting the cell-center-users (CCUs) and cell-edge-users (CEUs) for each pair. To enhance signal quality at the CEUs, we employ cooperative communication, where each CCU relays the inner-group common message to its paired CEU. This framework is formulated as an optimization problem by jointly optimizing user pairing, beamforming vectors at the base station (BS), common stream split, time slot duration, and transmit power of CCUs to maximize the network sum rate. The formulated problem is highly non-convex and difficult to solve, and hence, we adopt bi-level optimization which breaks the original problem into outer and inner problems. The outer problem is considered as the user pairing problem and we solve it using the SUS-MG algorithm. Once the users are paired, we solve the inner optimization problem for each pair using a successive convex approximation (SCA) approach. Finally, numerical results demonstrate that our proposed approach can outperform baseline schemes.