DRL for Trajectory Design and Resource Management in UAV-Aided RSMA Networks

This work explores the potential of Unmanned Aerial Vehicles (UAVs) to enhance traditional communication networks by leveraging their inherent mobility and adaptability. We address the problem of 3-dimensional flight design and resource management for a UAV-assisted network, where a UAV acts as an aerial base station employing rate-splitting multiple access to connect with ground devices. To achieve long-term sum rate maximization, we propose a method that uses deep reinforcement learning with a specific method called softmax deep double deterministic policy gradients (SD3). This approach optimizes transmission power, rate allocation, and UAV-BS trajectory parameters. Simulation results demonstrate that the SD3- based scheme converges better and achieves a higher sum rate than traditional DRL-based schemes.