Blind Separation Research Articles

Learning mid-IR emission spectra of PAH populations from observations

AbstractObservations of the mid-infrared (mid-IR, 3-15 μm) spectra of photo-dissociation regions reveal ubiquitous, broad and intense emission bands, the aromatic infrared bands (AIBs), attributed to polycyclic aromatic hydrocarbons (PAHs). Studies of the AIBs showed spectral variations (e.g. in the band positions) between different astrophysical objects, or even within single object, thanks to hyperspectral images. The James Webb Space Telescope (JWST) will allow to get further spectral and spatial details compared to former space observatories. This will come with large data sets, which will require specific tools in order to perform efficient scientific analysis.We propose in this study a method based on blind signal separation to reduce the analysis of such large data set to that of a small number of elementary spectra, spectrally representative of the data set and physically interpretable as the spectra of populations of mid-IR emitters. The robustness and fastness of the method are improved compared to former algorithms. It is tested on a ISO-SWS data set, which approaches the best the characteristics of JWST data, from which four elementary spectra are extracted, attributed to cationic, neutral PAHs, evaporating very small grains and large and ionized PAHs.

Identifying 2010 Xynthia Storm Signature in GNSS-R-Based Tide Records

In this study, three months of records (January–March 2010) that were acquired by a geodetic Global Navigation Satellite Systems (GNSS) station from the permanent network of RGP (Réseau GNSS Permanent), which was deployed by the French Geographic Institute (IGNF), located in Socoa, in the south of the Bay of Biscay, were used to determine the tide components and identify the signature of storms on the signal to noise ratio (SNR) during winter 2010. The Xynthia storm hit the French Atlantic coast on the 28th of February 2010, causing large floods and damages from the Gironde to the Loire estuaries. Blind separation of the tide components and of the storm signature was achieved while using both a singular spectrum analysis (SSA) and a continuous wavelet transform (CWT). A correlation of 0.98/0.97 and root mean square error (RMSE) of 0.21/0.28 m between the tide gauge records of Socoa and our estimates of the sea surface height (SSH) using the SSA and the CWT, respectively, were found. Correlations of 0.76 and 0.7 were also obtained between one of the modes from the SSA and atmospheric pressure from a meteorological station and a mode of the SSA. Particularly, a correlation reaches to 0.76 when using both the tide residual that is associated to surges and atmospheric pressure variation.

The CMB angular power spectrum via component separation: a study on Planck data

Aims. We investigate the extent to which foreground-cleaned cosmic microwave background (CMB) maps can be used to estimate cosmological parameters at small scales. Methods. We use the SMICA method, a blind separation technique that works directly at the spectral level. In this work we focus on the small scales of the CMB angular power spectrum, which are chiefly affected by noise and extragalactic foregrounds, such as point sources. We adapt SMICA to use only cross-spectra between data maps, thus avoiding the noise bias. In this study, performed using both simulations and Planck 2015 data, we fit for extragalactic point sources by modelling them as shot noise of two independent populations. Results. In simulations, we correctly recover the point-source emission law, and obtain a CMB angular power spectrum that has an average foreground residual of one fifth of the CMB power at ℓ ≥ 2200. With Planck data, the recovered point-source emission law corresponds to external estimates, with some offsets at the highest and lowest frequencies, possibly due to frequency decoherence of point sources. The CMB angular power spectrum residuals are consistent with what we find in simulations. The cosmological parameters obtained from the simulations and the data show offsets up to 1σ on average from their expected values. Biases on cosmological parameters in simulations represent the expected level of bias in Planck data. Conclusions. The results on cosmological parameters depend on the detail of the foreground residual contamination in the spectrum, and therefore a tailored modelling of the likelihood foreground model is required.

Blind separation of PCMA signals based on improved particle filter algorithm

For the problem of particle degradation and depletion in the blind separation of particle filter algorithm, particle swarm optimization algorithm and genetic algorithm are used to improve the traditional PCMA blind separation algorithm based on particle filter. The improved algorithm uses particle swarm optimization algorithm to optimize the particle and uses genetic algorithm to resample the particle. It can maintain the diversity of particles and reduce the problem of particle degradation and dilution effectively. The simulation shows that the improved blind separation algorithm has better separation performance.

Time-Domain Output Data Identification Model for Pipeline Flaw Detection Using Blind Source Separation Technique Complexity Pursuit

Vital defect information present in the magnetic field data of oil and gas pipelines can be perceived by developing such non-parametric algorithms that can extract modal features and performs structural assessment directly from the recorded signal data. This paper discusses such output-only modal identification method Complexity Pursuit (CP) based on blind signal separation. An application to the pipeline flaw detection is presented and it is shown that the complexity pursuit algorithm blindly estimates the modal parameters from the measured magnetic field signals. Numerical simulations for multi-degree of freedom systems show that the method can precisely identify the structural parameters. Experiments are performed first in a controlled laboratory environment secondly in real world, on pipeline magnetic field data, recorded using high precision magnetic field sensors. The measured structural responses are given as input to the blind source separation model where the complexity pursuit algorithm blindly extracted the least complex signals from the observed mixtures that were guaranteed to be source signals. The output power spectral densities calculated from the estimated modal responses exhibit rich physical interpretation of the pipeline structures.

Algorithms for Blind Separation and Estimation of Transmitter and Receiver IQ Imbalances

We propose a blind transmitter (Tx) and receiver (Rx) in-phase/quadrature (IQ) amplitude and phase imbalances estimation scheme based on a combination of analytical modeling and clustering algorithms. The proposed scheme can isolate and separately estimate Tx and Rx phase and amplitude imbalances, applicable to high-order multi-modulus signals, and hence, is a great physical-layer-monitoring tool for network disaggregation. We first estimate the Rx imbalances by identifying the fact that the received signal distribution in presence of laser frequency offset (FO), phase noise (PN), and Rx imbalance forms concentric ellipses. A convex hull assisted ellipse correction algorithm is used to determine the Rx IQ imbalance parameters that best match the ellipses. Standard laser FO compensation algorithm follows. A k -means clustering assisted blind phase search algorithm is then employed to mitigate the PN in presence of Tx imbalances. Finally, another iterative k -means clustering procedure is used to estimate the Tx IQ imbalances. The proposed scheme is evaluated experimentally for 34 Gbaud QPSK and 16-quadratic-amplitude modulation signals. The accurate estimation for a wide range of Tx and Rx IQ imbalances and good robustness to amplified spontaneous emission noise is demonstrated.

Underdetermined convolutive blind separation of sources integrating tensor factorization and expectation maximization

Underdetermined convolutive blind separation of sources is a challenging topic in speech and audio processing. In plenty of conventional separation algorithms, the source signal estimation is performed in the frequency-domain, thus leading to permutation problems and poor separation results. In this paper, to solve the permutation alignment and obtain better source separation results, we exploit the algebraic structure of tensor factorization model and expectation–maximization (EM) to provide a new time-frequency algorithm. This is because tensor factorization has an advantage to estimate the channel for the underdetermined convolutive mixture case, and EM algorithm is conducive to faster converge to the desired solution and better source separating property. In the proposed algorithm, we first estimate the mixing matrix by using tensor decomposition, while permutation alignment algorithm is used to deal with the permutation problems. Then, the model parameters are updated using EM algorithm for improving source separation performance. At the same time, the spatial images of source signals are obtained using Wiener filters constructed from the estimated parameters. Furthermore, the time-domain source signals can be obtained through inverse short-time Fourier transform. Finally, a series of simulations show that compared with some existing separation algorithms, the proposed algorithm achieves better separation performance.

Combining Spectral and Spatial Features for Deep Learning Based Blind Speaker Separation

This study tightly integrates complementary spectral and spatial features for deep learning based multi-channel speaker separation in reverberant environments. The key idea is to localize individual speakers so that an enhancement network can be trained on spatial as well as spectral features to extract the speaker from an estimated direction and with specific spectral structures. The spatial and spectral features are designed in a way such that the trained models are blind to the number of microphones and microphone geometry. To determine the direction of the speaker of interest, we identify time-frequency T-F units dominated by that speaker and only use them for direction estimation. The T-F unit level speaker dominance is determined by a two-channel chimera++ network, which combines deep clustering and permutation invariant training at the objective function level, and integrates spectral and interchannel phase patterns at the input feature level. In addition, T-F masking based beamforming is tightly integrated in the system by leveraging the magnitudes and phases produced by beamforming. Strong separation performance has been observed on reverberant talker-independent speaker separation, which separates reverberant speaker mixtures based on a random number of microphones arranged in arbitrary linear-array geometry.

A Blind Separation Method of PCMA Signals Based on MSGibbs Algorithm

A blind separation method of PCMA signals with different symbol rates based on MS-Gibbs (multiple states Gibbs) algorithm is proposed. The prior probability of input symbol pair is calculated based on channel state and two input signal components, which is used to guide the update of symbol sequence. Simulation results show that the performance of the proposed algorithm is similar to that of DG-PSP algorithm, but the complexity of separation is greatly reduced through the proposed algorithm. Compared with no iteration in the 10−2 order of magnitude of BER (bit error rate), the proposed algorithm can obtain nearly 2dB SNR (signal-noise ratio) gain after 2 iterations. If the number of iterations is 4, the proposed algorithm can obtain nearly 4dB SNR (signal-noise ratio) gain.

Automated classification of transient contamination in stationary acoustic data

An automated procedure for the classification of transient contamination of stationary acoustic data is proposed and analyzed. The procedure requires the assumption that the stationary acoustic data of interest can be modeled as a band-limited, Gaussian random process. It also requires that the transient contamination be of higher variance than the acoustic data of interest. When these assumptions are satisfied, it is a blind separation procedure, aside from the initial input specifying how to subdivide the time series of interest. No a priori threshold criterion is required. Simulation results show that for a sufficient number of blocks, the method performs well, as long as the occasional false positive or false negative is acceptable. The effectiveness of the procedure is demonstrated with an application to experimental wind tunnel acoustic test data which are contaminated by hydrodynamic gusts.

ABS: an analytical method of blind separation of CMB from foregrounds

Extracting CMB B-mode polarization from complicated foregrounds is a challenging task in searching for inflationary gravitational waves. We propose the ABS method as a blind and analytical solution to this problem. It applies to the measured cross bandpower between different frequency bands and obtains the CMB B-mode bandpower analytically. It does not rely on assumptions of foregrounds and does not require multiple parameter fitting. Testing against a variety of foregrounds, survey frequency configurations and instrument noise, we verify its applicability and numerical stability. The ABS method also applies to CMB temperature, E-mode polarization, the thermal Sunyaev–Zel’dovich effect, spectral distortion, and even significantly different problems such as cosmic magnification.

A Modified JadeR for Signal Separation under Gaussian Noise

A Modified version of Joint Approximation Diagonalization Estimation of Real Signals algorithm (JADER) is proposed to enhance efficiency and speed of Blind Signal Separation (BSS). MJADER based on the mixture's dimensions minimization step, where the cumulant matrices have been estimated using a reduced-dimension observed mixture. The approach (M-JADER) is based on a threshold step, it is easy to implement, computationally efficient and faster than standard JADER about 50% where it has less running time. The comparison done under tow types of niose(semi-white Gaussian noise and Uniform noise)

Underdetermined Reverberant Audio-Source Separation Through Improved Expectation–Maximization Algorithm

Underdetermined reverberant audio-source separation is an important issue in speech and audio processing. To solve this problem, many separation algorithms have been proposed, in which model parameter estimation is performed in the time–frequency domain, leading to permutation ambiguity and poor separation performance. Additionally, in the existing expectation–maximization (EM) algorithms, one of the crucial problem is that updating the model parameters at each iterative step is time-consuming. In this paper, we present an improved EM algorithm that combines nonnegative matrix factorization (NMF) and time differences of arrival (TDOA) estimation, avoiding the time consumption by properly selecting initial values of the EM algorithm. In the proposed algorithm, NMF source model is used to avoid the permutation ambiguity problem, and acoustic localization can be achieved by transforming the TDOA. Then, model parameters are updated to obtain better separation results. Finally, the source signals are separated using Wiener filters. The experimental results show that compared with existing blind separation methods, the proposed algorithm achieves better performance on source separation.

Minimum BER Criterion Based Robust Blind Separation for MIMO Systems

In this paper, a robust blind source separation (BSS) algorithm is investigated based on a new cost function for noise suppression. This new cost function is established according to the criterion of minimum bit error rate (BER) incorporated into maximum likelihood (ML) principle based independent component analysis (ICA). With the help of natural gradient search, the blind separation work is carried out through optimizing this constructed cost function. Simulation results and analysis corroborate that the proposed blind separation algorithm can realize better performance in speed of convergence and separation accuracy as opposed to the conventional ML-based BSS.

Constrained Blind Separation Algorithm Using Variable Step Size and Variable Momentum Factor

The traditional blind separation algorithm is mainly for the instantaneous mixing problem in the stable environment. In the practical applications, blind separation often takes into account the interference of the external environment, which requires that the algorithm has strong tracking performance, but the traditional algorithm can’t meet the needs. Aiming at the problem of instantaneous blind separation in non-stationary environment, constrained blind separation algorithm using variable step size and variable momentum factor is proposed in this paper. Based on the nonholonomic natural gradient algorithm, the cost function is constrained by the disturbance of the hybrid system and the constraint factors take the form of self-adaptive adjustment. According to the separation situation, the constraint factors are adjusted adaptively to accelerate the convergence speed. The variable step size based on the cost function gradient is introduced to improve the tracking performance. By incorporating momentum term, the momentum factor is adaptively adjusted to make it have better separation performance. The simulation results show that compared with the traditional algorithm, the proposed algorithm can better balance the contradiction between convergence speed and steady-state error in non-stationary environment, and has better separation performance. In the case of obvious disturbance in the mixed system, the algorithm can effectively improve the shortcomings of the traditional algorithm. In summary, constrained blind separation algorithm using variable step size and variable momentum factor proposed in this paper is effective.

TDOA estimation of dual-satellites interference localization based on blind separation

The time difference of arrival (TDOA) estimation plays a crucial role in the accurate localization of the satellite interference source. In the dual-satellites interference source localization system, the target signal from the adjacent satellite is likely to be interfered by the normal communication signal with the same frequency. Therefore, the signal to noise ratio (SNR) of the target signal would become too low, and the TDOA estimation through cross-correlation processing would be unreliable or even unattainable. This paper proposes a technique based on blind separation to solve the co-channel interference problem, where separation of the mixed signal can be carried out by the particle filter (PF) algorithm. The experimental results show that the proposed method could achieve more accurate TDOA estimation. The measured data obtained by using the software radio platform at 915 MHz and 2 GHz respectively verify the effectiveness of the proposed method.

Full-Duplex Cognitive Radio Using Guided Independent Component Analysis and Cumulant Criterion

Due to the ever-increasing wireless service requirements, the issues of a limited frequency resource and underutilized frequency allocation scheme as well as frequency crowd have come to become increasingly severe. It gives rise to being difficult to effectively eliminate the spectrum interference and fulfill an efficient spectrum access mechanism using the congested frequency band. For this purpose, this paper conceives a simultaneous transmission and sensing scheme based on the guided independent component analysis combined with the cumulant-based non-Gaussian criterion for fulfilling full-duplex cognitive radio (FD-CR). The philosophy of this proposed scheme is that the ICA based blind source separation principle is firstly utilized for blind separation of the observation mixtures, and then the separated signals are recognized through correlation processing and non-Gaussian criterion based on cumulant. In particular, the known secondary user signal is used as a signature signal for assisting ICA framework construction for blind extraction of signals. As a key component of machine learning, the ICA theory can contribute significantly to conquering the static sensing problem and residual self-interference influence in existing conventional cognitive radio systems. The simulation experiments and analysis corroborate the effectiveness of the proposed method and system design. In addition, the low complexity and performance enhancement of the proposed method are also validated by comparing with that of the two practical FD-CR schemes.

Enhanced NMF Separation of Mixed Signals in Strong Noise Environment

Separation of mixed signals from a noisy environment without prior conditions is one of the difficulties in blind signal separation. To solve the problem of poor separation effect of mixed signals in a strong noise environment, we propose an enhanced non-negative matrix factorization method in this paper. By extending the Kullback–Leibler divergence form, this method adopts a new target signal and noise estimation algorithm to overcome the shortcomings of existing methods in noise estimation. Furthermore, combining with the least squares algorithm, the computational complexity is effectively reduced, and the computational efficiency of the algorithm is improved while the source signals are well estimated. The theoretical analysis and simulation results show that the proposed algorithm is better than the existing algorithms in terms of the source signal separation from mixed signals with noise, especially when the signal and noise energy are equivalent and the mixed signals are completely obliterated in the noise, the proposed algorithm has more obvious advantages than the existing algorithms, while the operation efficiency has been improved.

Efficient blind nonparametric dependent signal extraction algorithm for determined and underdetermined mixtures

Blind extraction or separation statistically independent source signals from their linear mixtures have been well studied in the last two decades, which is realised by searching for local extrema of certain objective functions. In this paper, by employing nonparametric non-Gaussianity (NG) measure, a blind source separation/extraction (BSS/BSE) algorithm is derived to separate or extract statistically dependent source signals from their underdetermined or determined linear mixtures. Firstly, we show that maximisation NG measure can separate/extract statistically weak dependent source signals. Then, the nonparametric NG measure is defined by statistical distances between different cumulative distribution function (CDF) of separated signals, which can be estimated by quantiles and order statistics (OS) using L2 norm efficiently. Finally, the nonparametric NG measure aided algorithm is optimised by a deflation procedure. Simulation results for synthesis and real world data show that the proposed algorithm can extract the desired dependent source signals and yield expected performance.

A Source Counting Method Using Acoustic Vector Sensor Based on Sparse Modeling of DOA Histogram

The number of sources present in a mixture is crucial information often assumed to be known or detected by source counting. The existing methods for source counting in underdetermined blind speech separation suffer from the overlapping between sources with low W-disjoint orthogonality. To address this issue, we propose to fit the direction-of-arrival (DOA) histogram with multiple von-Mises density (VM) functions directly and form a sparse recovery problem, where all the source clusters and the sidelobes in the DOA histogram are fitted with VM functions of different spatial parameters. We also developed a formula to perform the source counting taking advantage of the values of the sparse source vector to reduce the influence of sidelobes. Experiments are carried out to evaluate the proposed source counting method, and the results show that the proposed method outperforms two well-known baseline methods.