Cluster Free Downlink Miso Noma System: Drl Approach

We propose a cluster-free downlink multiple-input-single-output (MISO) nonorthogonal multiple access (NOMA) system that leverages joint optimization of successive interference cancellation (SIC) and transmit beamforming for effective interference mitigation. Our system aims to maximize the overall system rate while satisfying quality of service requirements and SIC decision constraints. The joint optimization of transmit beamforming and SIC decisions in MISO NOMA systems presents an NP-hard problem, making traditional optimization approaches impractical. To address this challenge, we develop a dual deep reinforcement learning framework that combines the twin delayed deep deterministic (TD3) algorithm for transmit beamforming optimization with a deep Q network for SIC decision-making. Simulation results demonstrate that our proposed approach achieves near-optimal performance, closely matching the exhaustive search benchmark while maintaining computational efficiency. This work presents a significant advancement in the practical implementation of cluster-free NOMA systems, offering a scalable solution for next-generation wireless networks.