Acoustic Array Processing Research Articles

Pseudospectrum-based methods for estimating the wind speed and direction based on closely spaced microphone signals

Acoustic array processing can be employed to measure the wind speed and direction based on microphone signals. Turbulent pressure fluctuations picked up by microphones are referred to as wind noise. According to Taylor's frozen turbulence hypothesis, turbulent eddies retain their shape while advecting at nearly the mean wind speed and in the wind direction. It follows that wind noise propagates accordingly across a microphone array when the inter-microphone distance is smaller than the turbulence wavelength. This property can be exploited to track the orientation of the turbulence advection, and hence to characterize the wind flow. We propose beamforming and signal subspace-based methods to estimate the wind speed and direction using a compact microphone array. In particular, the pseudospectrum of measured wind noise is computed against candidate pairs of wind speed and direction. The wind speed and direction estimates are then obtained as the maximizers of the pseudospectrum. In addition, we extend an existing time difference of arrival-based method originally derived for three microphones to an arbitrary number of microphones. We evaluate the estimation accuracy of the proposed methods separately for the wind speed and direction. Possible applications include highly integrable, portable, and inexpensive anemometers for smart sensors or action cameras.

Generative models for sound field reconstruction.

This work examines the use of generative adversarial networks for reconstructing sound fields from experimental data. It is investigated whether generative models, which learn the underlying statistics of a given signal or process, can improve the spatio-temporal reconstruction of a sound field by extending its bandwidth. The problem is significant as acoustic array processing is naturally band limited by the spatial sampling of the sound field (due to the difficulty to satisfy the Nyquist criterion in space domain at high frequencies). In this study, the reconstruction of spatial room impulse responses in a conventional room is tested based on three different generative adversarial models. The results indicate that the models can improve the reconstruction, mostly by recovering some of the sound field energy that would otherwise be lost at high frequencies. There is an encouraging outlook in the use of statistical learning models to overcome the bandwidth limitations of acoustic sensor arrays. The approach can be of interest in other areas, such as computational acoustics, to alleviate the classical computational burden at high frequencies.

Enhancing fibre-optic distributed acoustic sensing capabilities with blind near-field array signal processing

Distributed acoustic sensors (DAS) can monitor mechanical vibrations along thousands independent locations using an optical fibre. The measured acoustic waveform highly varies along the sensing fibre due to the intrinsic uneven DAS longitudinal response and distortions originated during mechanical wave propagation. Here, we propose a fully blind method based on near-field acoustic array processing that considers the nonuniform response of DAS channels and can be used with any optical fibre positioning geometry having angular diversity. With no source and fibre location information, the method can reduce signal distortions and provide relevant signal-to-noise ratio enhancement through sparse beamforming spatial filtering. The method also allows the localisation of the two-dimensional spatial coordinates of acoustic sources, requiring no specific fibre installation design. The method offers distributed analysis capabilities of the entire acoustic field outside the sensing fibre, enabling DAS systems to characterise vibration sources placed in areas far from the optical fibre.

Statistically robust estimation of source bearing via minimizing the Bhattacharyya distance.

Source bearing estimation is a common technique in acoustic array processing. Many methods have been developed and most of them exploit some underlying statistical model. When applied to a practical system, the robustness to model mismatch is of major concern. Traditional adaptive methods, such as the minimum power distortionless response processor, are notoriously known for their sensitivity to model mismatch. In this paper, a parameter estimator is developed via the minimum Bhattacharyya distance estimator (MBDE), which provides a measure of the divergence between the assumed and true probability distributions and is, thus, capable of statistically matching. Under a Gaussian random signal model typical of source bearing estimation, the MBDE is derived in terms of the data-based and modeled covariance matrices without involving matrix inversion. The performance of the MBDE, regarding the robustness and resolution, is analyzed in comparison with some of the existing methods. A connection with the Weiss-Weinstein bound is also discussed, which gives the MBDE an interpretation of closely approaching a large-error performance bound. Theoretical analysis and simulations of bearing estimation using a uniform linear array show that the proposed method owns a considerable resolution comparable to an adaptive method while being robust against statistical mismatch, including covariance mismatch caused by snapshot deficiency and/or noise model mismatch.

Sullivan’s contributions to model-based acoustic array processing for underwater and beyond

Edmund Sullivan’s long and fruitful research efforts have spanned a number of fields, both underwater and ashore. Most notable to this author is his achievement on model-based processing for underwater acoustic arrays, documented in a 2015 Springer monograph. Through his unremitting dedication and long-time service to the Acoustical Society of America, and most prominently, the Technical Committee on Signal Processing in Acoustics, Ed’s achievements have long garnered utmost respect and reputability. This paper reviews his recent effort on model-based Bayesian signal processing with a rigorous handling of the Kalman filter, and acoustic array processing within the Bayesian framework. This paper also reflects personal remembrance of the mutually beneficial collaboration with Ed Sullivan in acoustic landmine detection and Bayesian analysis.

Sparse Representation Using Multidimensional Mixed-Norm Penalty With Application to Sound Field Decomposition

A sparse representation method for multidimensional signals is proposed. In generally used group-sparse representation algorithms, the sparsity is imposed only on a single dimension and the signals in the other dimensions are solved in the least-square-error sense. However, multidimensional signals can be sparse in multiple dimensions. For example, in acoustic array processing, in addition to the sparsity of the spatial distribution of acoustic sources, acoustic source signals will also be sparse in the time-frequency domain. We define a multidimensional mixed-norm penalty, which enables the sparsity to be controlled in each dimension. The majorization-minimization approach is applied to derive the optimization algorithm. The proposed algorithm has the advantages of a wide range for the sparsity-controlling parameters, a small cost of adjusting the balancing parameters, and a low computational cost compared with current methods. Numerical experiments indicate that the proposed method is also effective for application to sound field decomposition.

Compressive acoustic holography with block-sparse regularization.

Sparse reconstruction methods, such as Compressive Sensing, are powerful methods in acoustic array processing, as they make wideband reconstruction possible. However, when addressing sound fields that are not necessarily sparse (e.g., in acoustic near-fields, reflective environments, extended sources, etc.), the methods can lead to a poor reconstruction of the sound field. This study examines the use of sparse analysis priors to promote block-sparse solutions. In particular, a Fused Total Generalized Variation (F-TGV) method is developed, to analyze the sound field in the near-field of acoustic sources. The method promotes sparsity both on the spatial derivatives of the solution and on the solution itself, thus seeking solutions where the non-zero coefficients are grouped together. The performance of the method is examined numerically and experimentally, and compared with established methods. The results indicate that the F-TGV method is suitable to examine both compact and spatially extended sources. The method is promising for its generality, robustness to noise, and the capability to provide a wideband reconstruction of sound fields that are not necessarily sparse.

The correlation between the local environment and gray whale behavior as tracked using a sparse hydrophone array in Monterey Bay National Marine Sanctuary

Gray whales (Eschrichtius robustus) make one of the longest annual migrations of any marine mammal. While the seasonal periodicity of the migration and related travel corridors have been well-studied, little is known about how these whales use acoustics while migrating. Acoustic array processing is a powerful tool to localize and track vocalizing whales without disrupting their behavior. Concurrent with the NOAA Southwest Fisheries Science Center visual and infrared gray whale census, four hydrophone-recorder packages with sites separated by approximately 2 km in 58 to 110 m water in Monterey Bay National Marine Sanctuary offshore of Granite Canyon, California, recorded from November 2014 to June 2015. An automatic call detector identified gray whale M3 calls in the dataset. Spectrogram cross-correlation was used to measure the time-difference-of-arrival of these calls and estimate the location of the calling animal. These localizations provide gray whale swimming and calling behavior and indicate changes on diel and seasonal time scales. Results from 15 high-quality tracks show calling gray whales swimming at a mean speed of 1.97 m/s with a range of 1.18 to 2.54 m/s. Acoustic detections and tracks are compared with environmental parameters to investigate how gray whale behavior relates to their physical environment. These results are necessary for understanding the cues gray whales use to navigate from their feeding grounds offshore of Alaska to their breeding grounds in Baja California.

Sound source positioning using microphone array installed on a flying drone

This study presents an application of acoustic source localization using microphone array installed on a drone. Although recent technology of drone enables to record audio and video, there are still acoustic problems such as engine noise, propeller-induced flow noise, microphone directivity, etc. Such unwanted noise shows strong tonal characteristics with broad band noise contents. To acquire usable acoustic signal from array measurements while a drone is flying, we implemented acoustic signal conditioning and array processing. The approach to the problem is to adapt beamforming and TDOA techniques that provide localization capability of multiple sources and the position or direction of arrival of source, respectively. The array configuration and angle resolution are adjusted depending on the application requirements. We limit our attention to stable hovering condition to rule out position and heading error occurred by the drone motion. The experimental results of position estimation under a various noise conditions are discussed in the study.

The application of compressive sensing to underwater acoustic array processing and design

A brief overview of the application of compressive sensing to underwater acoustic array processing is given. Specific algorithms for estimating direction of arrival via L1 minimization are introduced. The Statistical Reduced Isometry Property (StRIP) is defined and used as a method to determine the ideal array designs for compressive beamforming. The tradeoff of using nested apertures and non-integer element spacing for improving StRIP over a broad range of frequencies is studied numerically. The array gain, which is a more directly useful performance metric, is also estimated using Monte Carlo methods. Compressive beamforming results from at-sea data taken on the Five Octave Research Array (FORA) will be presented. [Research funded by the Office of Naval Research.]

Initial development of wireless acoustic emission sensor Motes for civil infrastructure state monitoring

The structural state of a bridge is currently examined by visual inspection or by wired sensor techniques, which are relatively expensive, vulnerable to inclement conditions, and time consuming to undertake. In contrast, wireless sensor networks are easy to deploy and flexible in application so that the network can adjust to the individual structure. Different sensing techniques have been used with such networks, but the acoustic emission technique has rarely been utilized. With the use of acoustic emission (AE) techniques it is possible to detect internal structural damage, from cracks propagating during the routine use of a structure, e.g. breakage of prestressing wires. To date, AE data analysis techniques are not appropriate for the requirements of a wireless network due to the very exact time synchronization needed between multiple sensors, and power consumption issues. To unleash the power of the acoustic emission technique on large, extended structures, recording and local analysis techniques need better algorithms to handle and reduce the immense amount of data generated. Preliminary results from utilizing a new concept called Acoustic Emission Array Processing to locally reduce data to information are presented. Results show that the azimuthal location of a seismic source can be successfully identified, using an array of six to eight poor-quality AE sensors arranged in a circular array approximately 200 mm in diameter. AE beamforming only requires very fine time synchronization of the sensors within a single array, relative timing between sensors of 1 µs can easily be performed by a single Mote servicing the array. The method concentrates the essence of six to eight extended waveforms into a single value to be sent through the wireless network, resulting in power savings by avoiding extended radio transmission.

1P1-A15 自律型水中ロボットへの搭載を想定した小型3次元ソーナーの開発 : テストシステムの設計と基礎実験

To develop a small-size 3D sonar for Autonomous Underwater Vehicles, a test system, which consists by a T-type acoustic transducer array and signal processing units, is constructed. The signal processing units are controlled by a laptop PC, and can generate a special waveform to transmit, and record multi channel signals for post processing simultaneously. In this paper the design of the acoustic transducer array and the results of fundamental experiments in a test tank are described.

Investigation of an unusual noise phenomenon with horizontal and vertical hydrophone array data and three‐dimensional propagation modeling.

Acoustic data from an experiment conducted during Shallow Water 2006 (SW06) showed an unexplained broadband noise phenomenon. While the R/V Knorr towed a J‐15 acoustic source, which emitted a 93 Hz continuous‐wave signal, the R/V Endeavor performed Scanfish measurements in the vicinity to characterize internal wave phenomena. The WHOI Shark horizontal and vertical hydrophone array detected the 93 Hz signal, but eventually the signal was overpowered by broadband low‐frequency noise. The broadband noise may be associated with either research vessel but is uncharacteristically intense given that both vessels were greater than 25 km away from the Shark array. The complex acoustic environment due to the shelfbreak front, an internal wave packet, and the bathymetry of the continental shelf may have caused ducting of the noise field. Both three‐dimensional acoustic modeling and array processing techniques will be used to characterize the unexplained noise levels. In situ data from temperature sensors and Scanfish measurements provide environmental information needed to accurately model the sound speed field. These methods can also be applied to other SW06 experiments that were tailored to investigate the role of internal waves in acoustic propagation—such as those carried out on the R/V Sharp. [Work sponsored by the Office of Naval Research.]

Coherent acoustic array processing and localization on wireless sensor networks

Advances in microelectronics, array processing, and wireless networking have motivated the analysis and design of low-cost integrated sensing, computing, and communicating nodes capable of performing various demanding collaborative space-time processing tasks. In this paper, we consider the problem of coherent acoustic sensor array processing and localization on distributed wireless sensor networks. We first introduce some basic concepts of beamforming and localization for wide-band acoustic sources. A review of various known localization algorithms based on time-delay followed by least-squares estimations as well as the maximum-likelihood method is given. Issues related to practical implementation of coherent array processing, including the need for fine-grain time synchronization, are discussed. Then we describe the implementation of a Linux-based wireless networked acoustic sensor array testbed, utilizing commercially available iPAQs with built-in microphones, codecs, and microprocessors, plus wireless Ethernet cards, to perform acoustic source localization. Various field-measured results using two localization algorithms show the effectiveness of the proposed testbed. An extensive list of references related to this work is also included.

Signal models for broadband model-based array processors

The recently developed theory for broadband model-based acoustic array processing is based on a Kalman-type estimation scheme. Here two sets of equations are required; the state equations and the measurement equations. In the narrow-band case, the measurement equation is based on either a plane wave or curved wavefront monochromatic signal model. In the broadband case the signal model, i.e., the measurement equation, is represented as an autoregressive moving average (ARMA) process. In this approach the parameters of interest, which are the bearing and range, are combined with the ARMA parameters in a single state vector. This provides a scheme wherein these parameters are adaptively estimated in a recursive manner. However, the order of the ARMA process depends upon the spectral characteristics of the signal. In this study, various realistic broadband signal spectra are examined in terms of the required ARMA representations. Examples will be shown for several cases of representative surface ship radiation spectra.

Matched-beam processing: A synthesis of conventional beamforming and matched-field processing

Conventional beamforming, based on the concept of delay and sum of plane-wave signals, has been the cornerstone for radar and underwater acoustic array processing for many years. Matched-field processing, which has received much attention in recent years, represents a departure from conventional beamforming in that the signal propagation is an integral part of the processor. In this paper a framework is presented (i.e., matched beam processing) in which conventional beamforming and matched-field processing can be unified and treated on an equal footing. Matched-beam processing is matched-field processing in the beam domain. The resulting benefits are (1) the extension of full field processing to conventional beams, (2) the ability to do spatial filtering based on the angle of arrival, and (3) the ease of integration of the (matched-beam) processor into existing systems. For a horizontal array in shallow water, conventional beamforming would be the processor to use, but it suffers from signal gain degradation and bearing bias when the signal energy is split over many beams. Matched-field processing can be used to estimate the bearing accurately, but it must simultaneously estimate the target range, depth, and bearing. Using matched-beam processing, bearing can be estimated accurately while maintaining some tolerance to environmental mismatch.

Robust maximum energy adaptive matched field processing

Adaptive array processing algorithms have achieved widespread use because they are very effective at rejecting unwanted signals (i.e. controlling sidelobe levels) and in general have very good resolution (i.e. have narrow mainlobes). However, many adaptive high-resolution array processing algorithms suffer a significant degradation in performance in the presence of environmental mismatch. This sensitivity to environmental mismatch is of particular concern in problems such as acoustic array processing in the ocean where the array processor's knowledge of the propagation characteristics of the ocean is usually imperfect. A maximum energy matched field processor is formulated which tunes the replica vector used by a minimum variance distortionless response (MVDR) matched field processor to adjust for fluctuations in the environment. The resulting processor has the same structure as the two-stage MVDR processor interpretation of the adaptive minmax matched field processor proposed previously. However, the criteria which the two processors use to tune replica vector differ. The result is that the MEMV processor achieves a level of performance superior to that of the adaptive minmax processor and close to that of traditional adaptive processors while limiting the sensitivity of the processor to environmental mismatch. An algorithm for implementing the professor is developed, and the performance of the processor is analyzed. >

A minmax approach to adaptive matched field processing in an uncertain propagation environment

Adaptive array processing algorithms have achieved widespread use because they are very effective at rejecting unwanted signals (i.e., controlling sidelobe levels) and in general have very good resolution (i.e., have narrow mainlobes). However, many adaptive high-resolution array processing algorithms suffer a significant degradation in performance in the presence of environmental mismatch. This sensitivity to environmental mismatch is of particular concern in problems such as long-range acoustic array processing in the ocean where the array processor's knowledge of the propagation characteristics of the ocean is imperfect. An adaptive minmax matched field processor is formulated which combines adaptive matched field processing and minmax approximation techniques to achieve the effective interference rejection characteristic of adaptive processors, while limiting the sensitivity of the processor to environmental mismatch. An efficient implementation and alternative interpretation of the processor are developed. The performance of the processor is analyzed using numerical simulations. >

Nearfield high‐resolution source localization

An example of passive acoustic array processing consists of localizing the emanating machinery noise sources along a ship shell. When possible, an efficient way to increase the poor accuracy of this localization is to move the array Fresnel zone up to the sources. A linear beamforming with a spherical wave model then provides source bearing, range, and amplitude estimation. However, even for large antennas, the spatial resolution of this nearfield beamforming is limited to a wavelength fraction. To increase the localization accuracy, especially for low frequencies, “high‐resolution” methods should be used. This paper presents, in this case, Capon's method, also called adaptive beamforming. It's performances and resolving power are computed and validated through numerical simulations. They extend to the nearfield case the results usually found in a farfield situation. However, much more drastically than with plane waves, experimental results show that this method breaks down if the source model mismatches the wave fronts actually received. To avoid this strong model sensitivity, a robust approach, extending the spectral density matrix source subspace, is proposed. This method yields correct estimation of the amplitude and position of sources too close to be resolved by linear processing. This enhancement of Capon's method does not require an accurate source model. In addition, it has a low computational cost and is not nearfield dependent. [Work supported by DCN.]