Figure - available from: IET Communications
This content is subject to copyright. Terms and conditions apply.
Source publication
A dual‐path recurrent neural network model is proposed to achieve noise reduction of underwater acoustic signals, which consists of three steps: feature extraction, mask separation, and signal recovery. For feature extraction, we use a multi‐scale convolutional neural network to extract higher‐order non‐linear features of the input signal and chunk...
Similar publications
In the complex dynamic problems of practical engineering, the method of establishing dynamic equations and numerical analysis based on theories of mechanics is difficult to meet the needs of high dimension, multi-scale and high precision. In this paper, data-driven models of nonlinear dynamic systems are built through neural network. The numerical...
Citations
... Using the k-ϵ method, they applied viscous effects in the analysis and measured the propeller's acoustic power level and dimensionless coefficients. Y. Song et al. [17] presented a new technique to reduce underwater noise. They used a multi-scale conventional neural network and concluded that the applied model can improve the signal-to-noise ratio well. ...
Ship noise emission has many environmental impacts. Recognizing noise-generating source and investigating noise reduction techniques is the first step in elimination or reduction of these harmful effects. Studies show that ship propeller is one of the main sources of ship noise emission. Ducted propeller is one of the well-known solutions for noise reduction. In this paper, results of numerical solution for governing mathematical equations of noise emission in far field is presented and sound pressure level is calculated using Ffowcs Williams and Hawkings (FWH) equations. Numerical solution is validated by valid research resources. Propeller noise emission is also calculated by experimental measurements and filtering ambient noise frequencies. Finally, a comparison has been made between numerical solution and experimental results and the effect of duct on noise reduction is calculated. The measured SPL emitted from the propeller in CFD approach (solving the FWH) shows 28% reduction for ducted propeller in comparison with the propeller without the duct. In experimental approach (after removing the ambient noise) shows 37% reduction for ducted propeller in comparison with the propeller without the duct.
... Chirtu et al. [14] propose a new method for seismic signal denoising based on a U-Net CNN architecture, and use a publicly available dataset to prove the effectiveness of the proposed approach. Song et al. [15] propose a dual-path recurrent neural network model to achieve noise reduction in underwater acoustic signals, and experiments show that this model can improve the signalnoise ratio. For the problem of spectral reconstruction, Zhang et al. [16] propose a denoising autoencoder to reduce the noise in the original filter measurement before the reconstruction, proving that this method can obtain accurate reconstruction results when the measurement noise damages the data. ...
Gamma spectrum denoising can reduce the adverse effects of statistical fluctuations of radioactivity, gamma ray scattering, and electronic noise on the measured gamma spectrum. Traditional denoising methods are intricate and require analytical expertise in gamma spectrum analysis. This paper proposes a segmentation‐enhanced Convolutional Neural Network‐Stacked Denoising Autoencoder (CNN‐SDAE) method based on convolutional feature extraction network and stacked denoising autoencoder to achieve gamma spectrum denoising, which adopts the idea of data segmentation to enhance the learning ability of the neural network. By dividing the complete gamma spectrum into multiple segments and then using the segmentation‐enhanced CNN‐SDAE method for denoising, the method can achieve adaptive denoising without manually setting the threshold. The experimental results show that our method can effectively achieve gamma spectrum denoising while retaining the characteristics of the gamma spectrum. Compared with traditional methods, the denoising speed and effectiveness have been significantly improved, and the proposed method demonstrates an approximately 1.72‐fold enhancement in smoothing performance than the empirical mode decomposition method. Furthermore, in terms of retaining gamma spectrum characteristics, it also achieves a performance improvement of approximately three orders of magnitude than the wavelet method.
... In shallow water condition, underwater acoustics signals contain noise that originated from different sources such as reflection of surfaces, interference from fish or acoustics signal generated by ships. These facts explain that there are many techniques develop to reduce noise such as deep learning approach [1], value decomposition algorithm [2], wavelet transform [3], or neural networks [4]. These techniques are chosen to cope challenges in applying underwater acoustics signal for communications such as frequency-dependent attenuation [5], short range of communication [6], and very low bandwidth [7]. ...
... In general, Table I shows that noise increases the intensity of the source signal, which is about 3 dB, and this value is quite consistent across all signals with different distances. These results are certainly consistent with previous studies regarding intensity values before and after noise is removed [4]. With the variation in distance, the recorded signal attenuates as the distance between the transmitter and the receiver gets farther [9]. ...
With current developments, underwater communication using acoustic signals is widely used. Many things need to be prepared to support a reliable underwater communication system, such as taking measurements in a test tank to find out the correct measurement configuration. Underwater acoustic intensity measurements, which are detailed in this paper, are performed in the test tank using distance variation schemes. Measurements were made at various distances of 4, 10, 20, and 50 meters from the signal source. The hydrophone that was used has a sensitivity of -180 dB re 1V/µPa. The hydrophone was placed at a depth of 2 meters below the surface of the water in the test tank, which divided the test tank depth in half to ensure that reflections from the bottom and the surface were kept to a minimum. However, the problem is that there are noisy signals at different frequencies. This paper proposes a method using Noise Assisted - Multivariate Empirical Mode Decomposition (NA-MEMD) to decompose the signal and then calculate the sound intensity. The result shows that an increase in the distance between the transmitter and receiver, also causes a change in the intensity with an average change of 0.467 dB/meter. It is concluded that the NA-MEMD approach was shown to be successful in decomposing the intended signal from the noise to equalize the quality of the signal received at different distances, and the correlation between intensity value and change in distance is resilient, with a correlation value of 0.98, indicating a very strong correlation.
... Meanwhile, when the BER is greater than 0.01, the matrix tends to be a random matrix, the rank feature is not obvious, and the recognition rate tends to be 0. To summarize, due to the need for accurate scrambling-type information, the existing algorithms cannot achieve the full blind identification of linear block code scrambling. In recent years, deep learning methods have achieved good results in signal recognition [28,29], modulation recognition [30,31], malicious file detection [32,33], and other fields [34,35]. At the same time, deep-learningbased methods avoid manual feature extraction and have high practical value. ...
Aiming to solve the problem of the distinction of scrambled linear block codes, a method for identifying the scrambling types of linear block codes by combining correlation features and convolution long short-term memory neural networks is proposed in this paper. First, the cross-correlation characteristics of the scrambling sequence symbols are deduced, the partial autocorrelation function is constructed, the superiority of the partial autocorrelation function is determined by derivation, and the two are combined as the input correlation characteristics. A shallow network combining a convolutional neural network and LSTM is constructed; finally, the linear block code scrambled dataset is input into the network model, and the training and recognition test of the network is completed. The simulation results show that, compared with the traditional algorithm based on a multi-fractal spectrum, the proposed method can identify a synchronous scrambler, and the recognition accuracy is higher under a high bit error rate. Moreover, the method is suitable for classification under noise. The proposed method lays a foundation for future improvements in scrambler parameter identification.
