Deep Reinforcement Learning-Based Sum-Rate Maximization for Uplink Multi-user SIMO-RSMA Systems

This research aims to investigate a sum-rate maximization problem in uplink multi-user single-input multiple-output (SIMO) rate splitting multiple access (RSMA) systems. In these systems, Internet of Things devices (IoTDs) act as single-antenna nodes transmitting data to the multiple-antenna base station (BS) utilizing the RSMA technique. The optimization process includes determining the transmit powers of the IoTDs, decoding order, and receive beamforming vector at the BS. To achieve this goal, the problem is transformed into a deep reinforcement learning (DRL) framework, where a post-actor processing stage is proposed and a deep deterministic policy gradient (DDPG)-based approach is applied to tackle the issue. Via simulation results, we show that the proposed approach outperforms some benchmark schemes.