Energy Efficient Covert Transmission Strategy Based on Deep Reinforcement Learning

Investigation of the trade-off between transmission security and energy efficiency (EE) in wireless communication systems is of paramount importance. To address the energy efficient covert communication problem in non-orthogonal multiple access (NOMA) systems, a reinforcement learning approach—the deep deterministic policy gradient (DDPG)-based energy efficient covert communication algorithm—is proposed. The agent trained by the algorithm dynamically adjusts transmission strategies in each time slot based on channel state information (CSI). To better evaluate the performance of the proposed algorithm, the transmission strategy derived from a greedy algorithm is adopted as a benchmark. Simulation results demonstrate that compared to the Greedy strategy, the proposed algorithm achieves significant improvements in system EE while ensuring covertness in private message transmission.