This paper proposes a whistle-generative adversarial network (WhistleGAN) that generates whistles for biomimetic underwater covert acoustic communication. The proposed method generates new whistles to maintain covertness by avoiding the repetitive use of the same whistles. Since the human ear perceives octave frequency such that low-frequency resolution is relatively larger than that of low frequencies, the proposed WhistleGAN uses mel filter banks to keep the fidelity in mimicking while reducing the complexity. The mean opinion score test verified that the whistles generated by the proposed method and the recorded real whistles have a similar score of 4.3, and the computer simulations proved that the bit error rate performance of the proposed method is the same as that of the real whistle.

Underwater biomimetic communication (UBC) technology has been studied to overcome the restricted physical covertness of artificial acoustical communication signals. Recent advancements have demonstrated data transmission at 300 b/s with imperceptible mimicry levels and communication distances of up to 60 km. While these results have raised expectations for practical applications, the actual performance and maturity of this technology have remained unclear. This article investigates and evaluates existing biomimetic communication methods to clarify the current level of UBC technology to plan a development strategy. To analyze their performance, we classified them based on sound types, such as whistles and clicks, and modulation techniques. Simultaneously, the technological maturity of these methods is also assessed using the technical readiness level framework. The results show that biomimetic communication could be a more promising solution for military underwater communication requiring covertness. Finally, we suggest research directions to further develop this technology into a fully operational system.

Recently, the passive bistatic radar (PBR) which exploits frequency modulation (FM) radio transmitters as illuminators, has witnessed widespread usage owing to its various advantages. However, the characteristics of FM-radio-based PBR result in interference components in the range–Doppler (RD) map, which may increase false alarms. Therefore, this study proposes a method for suppressing interference components using a deep learning approach. The two main contributions of this study are as follows. First, a convolutional autoencoder model capable of effectively suppressing interference in the RD map of PBR was proposed. Second, a synthetic RD map dataset generation method that can enable the autoencoder to operate robustly in PBR in a real environment was presented. Further, a performance comparison between the proposed method and existing methods using simulated data proved that the deep learning-based method exhibited superior target detection performance. Furthermore, using the data recorded by the PBR in a real environment, the proposed autoencoder model was shown to effectively suppress interference components in a real interference environment.

Robust and continuous distance estimation of a rapidly approaching threat is a significant concern in underwater technology. Although frequency modulated continuous waves (FMCWs) have been widely used in near-range situations, FMCWs are challenging to use in a specific underwater situation. If a high-speed underwater target appears, the target echo can no longer be modeled as a simple Doppler-shifted version of the wave, which increases the distance estimation error of conventional FMCW methods. This situation also requires frequent updates of distance estimates. Thus, to obtain the distance of a high-speed underwater target, we propose a signal format of a frequency coded continuous wave (FCCW) and develop an FCCW-based distance estimation scheme. The proposed distance estimation algorithm has three steps: group correlators and accumulated sum (GCAS), time-delay map (TDM), and Doppler line fitting (DLF). The structure of the proposed FCCW enables the detection of each frequency hopping time (FHT) within a shorter period than the observation time. The GCAS algorithm detects FHTs in the FCCW and is designed to easily extend the matched filter length, resulting in an extension of the acquisition time. The TDM and DLF provide robust distance estimates and enable the classification of multiple targets. Simulation results show that the proposed method can detect FHTs and estimate the target distance more stably in low signal-to-noise-ratio environments by increasing the matched filter’s acquisition time.

This paper proposes a tractable and compact generative model for cetacean whistle signals based on Variational Autoencoder (VAE) and mixture of Gaussians in underwater biomimetic communication. Cetacean whistles are nonlinear frequency-modulated signals with diverse time-frequency structures and can be interpreted as probability density functions. To learn the marginal probability distribution of whistle data, this study incorporates a VAE and a mixture of Gaussians. The VAE learns latent space representations from whistle spectrogram images, and the prior distribution of the latent variables is modeled using multiple Gaussian components. Since the latent space of whistle contour is modeled as an explicit density function of multimodal structure, it enables unsupervised clustering and conditional whistle generation by analyzing component-wise features. The statistical distance between the synthesized whistles and the training data was evaluated using the Fréchet inception distance, Recall, Precision, and F1-score yielding a score of 2.6767, 87.21%, 91.99%, and 89.54%, respectively, when the number of Gaussian components was set to 10,000.

This study proposes a Variational AutoEncoder (VAE)-based Quality Controlled (QC) generative augmentation framework for North Atlantic right whale (NARW) upcall detection. Existing generative augmentation methods can generate synthetic samples; however, they do not provide a sample-level criterion for determining whether each generated sample is a positive sample that contributes to improved detector performance or a synthetic outlier that should be removed. This study learns manually extracted upcall frequency contours using a 1D-VAE and evaluates generated contour candidates in a 10-dimensional acoustic morphology feature space. The QC score is computed with respect to the reference distribution of manually extracted real upcall contours, and stochastic acceptance probabilities for borderline samples around the hard threshold are calibrated using same-call manual re-extraction variability. QC-passed contours are converted into detector training spectrograms using smooth amplitude modulation and Gaussian noise injection based on SNR statistics. Using 30,000 acoustic segments from the Kaggle NARW dataset, Original, Denoising Diffusion Probabilistic Models (DDPM), Contour-VAE without QC, and VAE-QC conditions were compared under the same detector and augmentation budget. VAE-QC with α = 0.97 achieved the highest mean AUC of 0.901, outperforming Original training (0.802), DDPM (0.845), and Contour-VAE without QC (0.810). Feature distribution and QC-score analyses further showed that VAE-QC suppresses morphology outlier tails observed in unfiltered generation. These results indicate that the key factor in generative augmentation is the QC process that defines feature boundaries useful for detector learning and selects synthetic positive samples accordingly.