Optimization Methods for Covert Communication Strategies with Unknown Transmit Power to the Warden in Underwater Acoustic Communication

This paper presents an optimization framework for underwater acoustic covert communication strategies with unknown transmit power at the warden, aiming to improve covert transmission performance in ocean Internet of Things (OIoT) applications. Existing UACC studies usually rely on the assumption that the warden has prior knowledge of the transmit power parameter, which enables standard detection methods, such as the likelihood ratio test (LRT), to be applied effectively. However, the warden often lacks prior knowledge in practical underwater acoustic (UWA) scenarios. To address this problem, a GLRT-based detection framework is developed for the case where Alice’s transmit power is unknown to the warden. Specifically, the unknown transmit power is estimated by MLE and then incorporated into the detection statistic. The convergence behavior of the transmit-power estimator is analyzed, and the corresponding detection error probability at the warden is derived. Based on this analysis, the covert throughput is maximized by jointly optimizing Alice’s transmit power and the number of transmission slots. Simulation results demonstrate that the proposed GLRT detector consistently achieves a lower detection error probability than both the mismatched LRT baseline and the UWA radiometer detector under transmit-power uncertainty. At a transmit power of 22 dBW, the proposed GLRT reduces the detection error probability by approximately 53% compared with the mismatched LRT baseline and by approximately 14% compared with the UWA radiometer detector.