Single Channel Blind Source Separation Research Articles

Single-Channel Blind Source Separation in Wireless Communications: A Complex-Domain Deep Learning Approach

Single-Channel Blind Source Separation in Wireless Communications: A Complex-Domain Deep Learning Approach

A single-channel blind source separation algorithm based on improved wavelet packet and variational mode decomposition

A single-channel blind source separation algorithm based on improved wavelet packet and variational mode decomposition

Single channel blind source separation of rolling bearing compound faults based on self-learning sparse decomposition and feature mode decomposition

Feature mode decomposition (FMD) has advantages over the other newer time-frequency methods such as ensemble empirical mode decomposition (EEMD) and variational mode decomposition (VMD) in single channel blind source separation (SCBSS). However, FMD has the defect of needing to determine the precise number of fault sources manually. To solve the above defect of FMD, an adaptive method for determining the number of fault sources based on the shift invariant sparse code (SISC) is proposed. SISC was used to train a set of basis functions from the single channel signal, and the corresponding potential components were reconstructed firstly. Subsequently, the structural similarity of these potential components was used for clustering, and each of the obtained clustering signals represented one kind of fault. Then the number of clustering was determined by minimizing the structural correlation among the clustering signals. It was considered that the source separation had achieved the best effect when the structural difference among the clusters was the largest, and the number of clustering at this time was used as the optimal estimated value, which was used as the modal inputs number of FMD calculation model to realize SCBSS of rolling bearing. Simulation and experimental analysis were carried out to verify the effectiveness of the proposed method, and its superiority was also verified through comparison.

Spaceborne Algorithm for Recognizing Lightning Whistler Recorded by an Electric Field Detector Onboard the CSES Satellite

The electric field detector of the CSES satellite has captured a vast number of lightning whistler events. To recognize them effectively from the massive amount of electric field detector data, a recognition algorithm based on speech technology has attracted attention. However, this approach has failed to recognize the lightning whistler events which are contaminated by other low-frequency electromagnetic disturbances. To overcome this limitation, we apply the single-channel blind source separation method and audio recognition approach to develop a novel model, which consists of two stages. (1) The training stage: Firstly, we preprocess the electric field detector wave data into the audio fragment. Then, for each audio fragment, mel-frequency cepstral coefficients are extracted and input into the long short-term memory network for training the novel lightning whistler recognition model. (2) The inference stage: Firstly, we process each audio fragment with the single-channel blind source to generate two different sub-signals. Then, for each sub-signal, the mel-frequency cepstral coefficient features are extracted and input into the lightning whistler recognition model to recognize the lightning whistler. Finally, the two results above are processed by decision fusion to obtain the final recognition result. Experimental results based on the electric field detector data of the CSES satellite demonstrate the effectiveness of the algorithm. Compared with classical methods, the accuracy, recall, and F1-score of this algorithm can be increased by 17%, 62.2%, and 50%, respectively. However, the time cost only increases by 0.41 s.

The Single-channel blind source separation based on VMD and Tukey M estimation for rolling bearing composite fault diagnosis

Rolling bearing is one of the core components in rotating machinery, and its running status directly affects the operation of the whole equipment. Faults of rolling bearings in the actual working process are often multiple faults. To effectively separate fault sources, the blind source separation method is used for the compound fault diagnosis of rolling bearings. Because of the impact of the number of artificially limited decompositions and quadratic penalty factor on VMD in the decomposition process, and the slow convergence and low accuracy in the objective function of traditional FastICA operation, the VMD algorithm based on the energy loss coefficient and the information entropy is proposed, which adaptively determines the number of modal components and the quadratic penalty factor; The Tukey M estimation is selected as the objective convergence function of the FastICA algorithm to improve its robustness. First, VMD is used to decompose the signal; Secondly, the original signal and the decomposed IMF component are reconstructed, the covariance matrix and the singular value decomposition are constructed, the number of fault sources is estimated by the proximity dominance method, and the decomposed IMF components are filtered through correlation analysis and kurtosis index to build a multi-channel feature set; Finally, the constructed multi-channel feature set is input to the FastICA algorithm based on the Tukey M estimation for the separation of fault source signals to achieve composite fault diagnosis. The compound fault experiment shows that the proposed method in this paper can effectively realize the blind source separation of rolling bearing fault features to realize the compound fault diagnosis in different positions.

An Evaluation of Non-Contact Photoplethysmography-Based Methods for Remote Respiratory Rate Estimation

The respiration rate (RR) is one of the physiological signals deserving monitoring for assessing human health and emotional states. However, traditional devices, such as the respiration belt to be worn around the chest, are not always a feasible solution (e.g., telemedicine, device discomfort). Recently, novel approaches have been proposed aiming at estimating RR in a less invasive yet reliable way, requiring the acquisition and processing of contact or remote Photoplethysmography (contact reference and remote-PPG, respectively). The aim of this paper is to address the lack of systematic evaluation of proposed methods on publicly available datasets, which currently impedes a fair comparison among them. In particular, we evaluate two prominent families of PPG processing methods estimating Respiratory Induced Variations (RIVs): the first encompasses methods based on the direct extraction of morphological features concerning the RR; and the second group includes methods modeling respiratory artifacts adopting, in the most promising cases, single-channel blind source separation. Extensive experiments have been carried out on the public BP4D+ dataset, showing that the morphological estimation of RIVs is more reliable than those produced by a single-channel blind source separation method (both in contact and remote testing phases), as well as in comparison with a representative state-of-the-art Deep Learning-based approach for remote respiratory information estimation.

Single-channel blind source separation of underwater acoustic signals using improved NMF and FastICA

When automatic monitoring buoys receive mixed acoustic signals from multiple underwater acoustic targets, the statistical blind source separation (BSS) task is used to separate the signals and identify vessel features, which is overly complex and needs improvement, especially noting that noise cancellation and stealth technologies are advancing rapidly. To fill this gap in capability, an improved non-negative matrix factorization (NMF) based BSS algorithm is built on a FastICA machine learning backbone. With this tool, the spatial and spectral correlation of underwater acoustic signals is introduced into the NMF algorithm improved by to resolve the non-convex and feature correlation problems commonly encountered by contemporary NMF algorithms. Moreover, the improved modulation feature adaptability of the NMF increases the local expressivity and independence of the decomposed base matrix, which is proven to meet the requirements of FastICA and used to improve the BSS effect of the FastICA. Simulated and empirical results show that compared with state-of-the-art FastICA and NMF based BSS algorithms, our novel approach obtains better signal-to-noise reduction and separation accuracy while maintaining superior target signal recognition features.

Single channel convolutive blind source separation for LFM radar signals

Abstract We propose a single channel blind source separation algorithm for convolutively mixed linear frequency modulation (LFM) signals based on smoothed Wigner-Ville distribution (SWVD) time-frequency analysis, Canny edge detection, and Hough transform detection. First, the SWVD time-frequency analysis diagram is obtained as an image based on the LFM time-frequency characteristics. Second, Canny edge detection is performed on the image. Then, Hough transform is used to detect the characteristic parameters of the linear signal. Finally, the source signal is recovered. The simulation results show that the algorithm is effective for single channel detection and extraction of convolutively mixed LFM signals.

Single-Channel Blind Separation of Unsynchronized Multiuser PSK Signals With Non-Identical Sampling Frequency Offsets

Single-channel blind source separation (SCBSS) of uncoordinated, non-spread, co-frequency interfering multi-user signals is a challenging task. When the multi-user signals have sampling frequency offset (SFO) due to the imperfection of the local oscillator (LO), this problem becomes even more difficult to be solved. In this letter, a select-stitch data-aided channel estimation (CE) initialized per-survivor processing (SS-DCE-PSP) algorithm is proposed for SCBSS of multi-user signals with SFOs caused by imperfect LOs. By dividing a long sequence into a number of segments and selecting the segments with highest detection reliability to stitch together, the time slip and symbol slip due to the non-identical SFOs can be resolved. The data-aided CE in SS-DCE-PSP further enhances the performance of PSP, especially for the initial symbols. Simulation results show that the proposed SS-DCE-PSP algorithm is robust to the normalized SFO up to 10−4, and the BERs obtained by the SS-DCE-PSP algorithm are close (within 0.5dB) to the BERs obtained for multi-user signals without SFOs.

Single channel blind source separation of complex signals based on spatial‐temporal fusion deep learning

Blind Source Separation (BSS) of complex signals composed of radar, communication and jamming signals is the first step in an integrated electronic system, which requires higher accuracy of separation. However, the traditional Single-Channel Blind Source Separation (SCBSS) method has low separation accuracy and poor robustness. Aiming at this problem, this paper proposes a SCBSS method based on spatial-temporal fusion deep learning model. This is a deep neural network model, which realizes spatial-temporal of mixed signals by integrating Convolutional Neural Network (CNN) and Bidirectional Long Short-Term Memory Network (BiLSTM). Convolutional Neural Network is used to extract spatial features from input sequences, and BiLSTM is used to mine timing rules of signals. A batch normalisation layer and a dropout layer are added to improve stability and prevent overfitting. The experiments show that the average similarity coefficient of the separated signals is above 0.99 and the Signal-Distortion Ratio (SDR) is up to 27 dB without noise. When the Signal-Noise Ratio is 0–20 dB and Jamming-Signal Ratio is 15 dB, the SDR is 5–30 dB higher than the traditional methods and the single network structure deep learning methods.

Heart-lung sound separation by nonnegative matrix factorization and deep learning

Auscultation is an ordinary way to diagnose cardiovascular and respiratory diseases. The heart-lung sound (HLS) from the stethoscope is a mixture of heart sound (HS) and lung sound (LS). The first and most crucial step is to differentiate between HS and LS from the mixture signal. This paper proposes an unsupervised single-channel blind source separation (BSS) algorithm based on nonnegative matrix factorization (NMF) and deep learning (DL). Firstly, HS signals are extracted by the multi-constrained NMF (MCNMF) decomposition algorithm and K-means & SVM (KS) clustering method proposed in this paper, and different types of HS signals are classified by a convolutional neural network (CNN). Then, the HLS is separated by the embedding space centroid network (ECNet) proposed in our study. The method outperforms the known state-of-the-art techniques in all evaluation metrics SNR, SDR, SIR, SAR, and correlation coefficient. Our algorithm has two advantages: 1) this study is the first to combine NMF and DL for the HLS separation. MCNMF with selected constraints can have a more substantial sparsity effect that improves the decomposition accuracy. KS clustering algorithm can distinguish the HS component from the LS component in two clustering results. By mapping classified HLS into an embedding space, the ECNet demonstrates excellent separation results. 2) HS extraction and classification first, then HLS separation. Because our method exploits more detailed information from HS types in HLS before BSS, the ECNet can outperform other HLS separation methods.

One to multiple mapping dual learning: Learning multiple signals from one mixture

Single channel blind source separation (SCBSS) refers to separating multiple sources from a mixture collected by a single sensor. Existing methods for SCBSS have limited performance in separating multiple sources and generalization. To address these problems, an algorithm is proposed in this paper to separate multiple sources from a mixture by designing a parallel dual generative adversarial network (PDualGAN) that can build the relationship between a mixture and the corresponding multiple sources to achieve one-to-multiple cross-domain mapping. This algorithm can be applied to a variety of mixtures including both instantaneous and convolutive mixtures. In addition, new datasets for single channel source separation are created which include the mixtures and corresponding sources for this study. Experiments were performed on four different datasets including both one-dimensional and two-dimensional signals. Experimental results show that the proposed algorithm outperforms state-of-the-art algorithms, measured with peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), source-to-distortion ratio (SDR), source-to-interferences ratio (SIR), relative root mean squared error (RRMSE) and correlation.

A Blind Filtering Framework for Noisy Neonatal Chest Sounds

Chest sound— as the first and most commonly available vital signal for newborns— contains affluent information about their cardiac and respiratory health. However, neonatal lung sound auscultation is currently challenging and often unreliable due to the noise and interference, particularly for preterm infants. The noise often overlaps with the heart and lung contents in both time and frequency. Moreover, the frequency band of the useful components varies from one case to another, making it difficult to separate by fixed band-pass filtering. In this study, a single-channel Blind Source Separation (SCBSS) framework is proposed to separate newborns’ lung and heart sounds from noisy chest sounds recorded by a digital stethoscope. This method first decomposes the signal into a multi-resolution representation using a time-frequency transform, and then applies source separation algorithms, to find proper ad hoc frequency filters. In the simulation scenario, two different time-frequency transforms are considered; Stationary Wavelet Transform (SWT) with dyadic bases, and Continuous Wavelet Transform (CWT) with redundant bases. The transforms are followed by three different source separation methods, namely Principal Component Analysis (PCA), Periodic Component Analysis (πCA), and Second Order Blind Identification (SOBI). The yielded combinations are applied to the chest sounds recorded from ninety-one preterm and full-term newborns. The results show that compared to raw signals, fixed band-pass filtering and seven other separation methods, the heart and lung sounds extracted by the proposed methods have higher quality index and also result in more reliable heart and respiratory rate estimation.

Mechanical Compound Fault Analysis Method Based on Shift Invariant Dictionary Learning and Improved FastICA Algorithm

For mechanical compound fault, it is of great significance to employ the vibration signal of a single-channel compound fault to analyze and realize the separation of multiple fault sources, which is essentially the problem of single-channel blind source separation. Shift invariant K-means singular value decomposition (shift invariant K-SVD) dictionary learning is suitable to extract the periodic and repeated fault features of a rotating machinery fault, hence in this article a single-channel compound fault analysis method is put forward which combines shift invariant K-SVD with improved fast independent component analysis (improved FastICA) algorithm. Firstly, based on single-channel compound fault signal, the shift invariant K-SVD algorithm can be used for learning multiple latent components that can be constructed as a virtual multi-channel signal. Then the improved FastICA algorithm is utilized to realize the separation of multiple fault source signals. With regard to the FastICA algorithm, the third-order convergence Newton iteration method is adopted to improve convergence speed. Moreover, in order to address the problem that FastICA is very sensitive to initialization, a steepest descent method can be applied. The experimental analysis of the compound fault of rolling bearing verifies that the presented method is effective to separate multiple fault source signals and the improved FastICA algorithm can increase convergence rate and overcome the problem of sensitivity to initialization.

Single-Channel Blind Source Separation of Spatial Aliasing Signal Based on Stacked-LSTM.

Aiming at the problem of insufficient separation accuracy of aliased signals in space Internet satellite-ground communication scenarios, a stacked long short-term memory network (Stacked-LSTM) separation method based on deep learning is proposed. First, the coding feature representation of the mixed signal is extracted. Then, the long sequence input is divided into smaller blocks through the Stacked-LSTM network with the attention mechanism of the SE module, and the deep feature mask of the source signal is trained to obtain the Hadamard product of the mask of each source and the coding feature of the mixed signal, which is the encoding feature representation of the source signal. Finally, characteristics of the source signal is decoded by 1-D convolution to to obtain the original waveform. The negative scale-invariant source-to-noise ratio (SISNR) is used as the loss function of network training, that is, the evaluation index of single-channel blind source separation performance. The results show that in the single-channel separation of spatially aliased signals, the Stacked-LSTM method improves SISNR by 10.09∼38.17 dB compared with the two classic separation algorithms of ICA and NMF and the three deep learning separation methods of TasNet, Conv-TasNet and Wave-U-Net. The Stacked-LSTM method has better separation accuracy and noise robustness.

A New Method for Separating EMI Signal Based on CEEMDAN and ICA

Aiming at the problem of electromagnetic interference signal separation, we propose a single-channel blind source separation method based on complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and independent component analysis (ICA). Firstly, decompose the mixed interference signal by CEEMDAN to obtain a series of intrinsic mode functions (IMF). Secondly, combine them into a new multi-dimensional signal and solve its correlation matrix, and use singular value decomposition to obtain the eigenvalues of the matrix, and then use it to estimate the source number. Thirdly, select those IMFs whose correlation coefficients with the observed signal are bigger and regard them together with the original signal after denoising as new observed signals. Finally, recover the source signals by fast independent component analysis (Fast-ICA). We also extend our method to multi-channel BBS, and experiment results show that our method can eliminate unnecessary noise in the signal and perform well.

Sequence to point learning based on bidirectional dilated residual network for non-intrusive load monitoring

Non-Intrusive Load Monitoring (NILM) or Energy Disaggregation, seeks to save energy by decomposing corresponding appliances power reading from an aggregate power reading of the whole house. It is regarded as a single channel blind source separation problem to extract sources from a mixed signal. Recent studies have shown that deep learning is widely applied to NILM problem. Theoretically,the ability of any neural network to extract load features is closely related to its depth. However, a deep neural network is difficult to train because of exploding gradient, vanishing gradient, and network degradation. Therefore, Bi-TCN residual block, inspired by a temporal convolution network (TCN), is applied to solve these problems. Causal dilated convolution is replaced by bidirectional(non-causal) dilated convolution to enlarge the receptive field of network and improve the performance of model. Two forms of residual connections are introduced to deep models. One is designed to facilitate training deep models, and the other is pursuing performance-boosting by combining load features extracted of different hierarchical levels to final prediction. We propose a sequence to point learning based on bidirectional dilated convolution for NILM on low-frequency data, called BitcnNILM. We compare our method with existing algorithms on low-frequency data via REDD and UK-DALE datasets. Experiments show that the superiority of our BitcnNILM in both load disaggregation and load on/off identification.

Sequence-to-Point Learning Based on Temporal Convolutional Networks for Nonintrusive Load Monitoring

Nonintrusive load monitoring (NILM) is performed by monitoring the total electricity consumption data from the customer’s meter and decomposing it with a load decomposition algorithm to get real-time electricity consumption data for individual appliances. Nonintrusive load decomposition is a single-channel blind source separation problem, which is difficult to implement because of the inherently unidentifiable problem. Recent studies have shown that deep learning is gradually becoming the mainstream approach to solve the NILM problem owing to the availability of data, computing power, and deep network training algorithm models. Among them, the sequence-to-point (seq2point) (Zhang) load decomposition model achieves state-of-the-art prediction because it concentrates on the representation ability of the network at one point. In this article, we optimize the structure of the network model based on the seq2point model, use a temporal convolutional network (TCN) with a more flexible perceptual field of view to train the load decomposition model, and replace the traditional activation function Relu with the activation function GELU. Finally, we test the algorithm in the public dataset UK-DALE and select the mean absolute error and F-score as the main evaluation indexes of the algorithm performance. Comparing the decomposition results of the two algorithms for the same period dataset, the results show that the temporal convolutional network model has substantially improved the performance index of load decomposition while reducing the model parameters.

SCBSS Signal De-Noising Method of Integrating EEMD and ESMD for Dynamic Deflection of Bridges Using GBSAR

Ground-based synthetic aperture radar (GBSAR) technology, as a new measurement technology, has the advantages of noncontact measurements, high precision, and all-weather measurement capability, and it has been widely used for bridge dynamic deflection measurements. In order to reduce the influence of noise in dynamic deflection of bridges obtained using GBSAR, this article proposes a single-channel blind source separation signal (SCBSS) de-noising method to obtain the denoised dynamic deflection of bridges. First, the extreme-point symmetric mode decomposition (ESMD) method and the ensemble empirical mode decomposition (EEMD) method are used to decompose the obtained dynamic deflection—as the original observation signal—into a series of intrinsic mode functions (IMFs) and a residual R. Second, the Spearman's Rho of each IMF with the original observation signal is calculated to remove the dominant IMFs of high-frequency noise. Third, the remaining IMFs and R decomposed by ESMD and EEMD are reconstructed into two sets of new signals, which form a new virtual multichannel data with the original observation signal. Finally, blind source separation is performed on the new virtual multichannel signal to obtain separated signal components. The separate signal components are converted in the frequency domain using the fast Fourier transform algorithm, and the noise signal components are identified using a spectrum analysis, to achieve further removal of noise information. The results of both simulated and on-site experiments show that the SCBSS signal de-noising method has a powerful signal de-noising ability.

Single-channel blind source separation based on attentional generative adversarial network

Blind single-channel source separation is a long-standing machine learning and signal processing problem. Traditional blind source separation (BSS) algorithms were proposed to solve this task utilizing multiple signal constraints. Generative adversarial network (GAN) are free from statistical constraints and samples, but the role of adversarial training in the task of BSS has not been fully demonstrated. Therefore, a new separation network model that enables to learn the known separated signal distribution from the stepwise fine estimation of the unknown mixture distribution was presented in this paper, and a self-attention mechanism was introduced to solve the problem of the blurring details of the generated image by the generator which preserves image details in the process of image separation. Compared with the existing single-channel blind source separation algorithm based on generative adversarial network-neural egg separation (NES), the detailed information of this new separation algorithm is more prominent, and the source signal in the mixed image has been separated more effectively, and has better separation performance than the classic blind source separation algorithms.