Deep Reinforcement Learning Based Beamforming Design in UAV-enabled NOMA ISAC System

In this paper, we investigate an unmanned aerial vehicle (UAV) enabled non-orthogonal multiple access (NOMA) empowered integrated sensing and communication (ISAC) system, where the UAV is introduced as the ISAC base station providing services to multiple users employing NOMA, while communication signals is simultaneously exploited for target sensing. To maximum the communication sum-rate while satisfying the maximum Cramér-Rao lower bound (CRLB) requirement, a beamforming design problem is formulated. The optimization problem is inherently non-convex, which makes it challenging to find the optimal solution through traditional optimization algorithms. To address this challenge, we adopt an online decision-making framework based on deep reinforcement learning (DRL) to solve the problem. Numerical results indicate that the proposed NOMA-ISAC outperforms conventional ISAC in terms of sum-rate.