Pair Of Microphones Research Articles

Time delay estimation of acoustic tones with unpredictable frequency variations under low signal to noise ratio conditions

This paper describes a novel approach for autonomously detecting and tracking aircraft in the vicinity of a unmanned aerial vehicle (UAV). The time difference of arrival (TDOA) of acoustic tones emanating from the distant aircraft are correlated between spatially distributed microphone pairs located on the UAV. The geometry of multiple microphone pairs then allows the elevation and azimuth of the approaching aircraft to be estimated, despite the high levels of onboard narrow- and broadband noise emanating from the engine firing sequence, propeller, airflow over the microphones, and mechanical vibration. Current signal processing techniques estimate time delay when the signal level of the approaching aircraft is above the background signature of the sensing aircraft, despite subtle frequency variations in the narrowband components of the approaching aircraft’s signature. The technique described in this paper is designed to track these unpredictable frequency variations and thus extends the detection range of the approach by enabling estimation of TDOA when the signal level is 20 dB below the noise floor. Potential detection ranges in excess of 1 km are demonstrated when the signal processing is combined with careful suppression of the many noise sources onboard the UAV.

Subspace-Based Identification of Acoustic Noise Spectra in Induction Motors

In this paper, we study the identification of acoustic noise spectra in induction motors by using a recently developed frequency-domain cross-power spectrum estimation algorithm. This algorithm is a noniterative high-resolution spectral estimator. In a test rig, from multiple experiments sound data are collected by an array of five-microphones placed hemispherically around motors in a reverberant and noisy room. In order to explore the issue of assembly micromisalignments, each motor is removed from the test rig and then replaced, after which the experiment is then repeated. The identification algorithm is used to detect changes in acoustic noise spectra of induction motors due to mechanical and electrical faults most frequently encountered in industry. Not only the autopower spectra of the individual microphones, but also the cross-power spectra of the microphone pairs are estimated. As a byproduct, it is demonstrated that one microphone is sufficient to identify noise spectra. The estimated acoustic spectra, or more compactly statistics extracted from them, can be used in the development of preventive maintenance programs for induction motors in service.

Two-stage Localisation Scheme Using a Small-scale Linear Microphone Array for Indoor Environments

The small-scale linear microphone arrays that are widely found in smartphones could be used to locate a sound source for indoor environments. After the Time Differences Of Arrival (TDOAs) in microphone pairs are estimated, a TDOA-based hybrid localisation scheme is proposed for a small-scale linear microphone array. The scheme contains two stages: the initialisation stage using the Levenberg-Marquardt (LM) algorithm, and the refining solution stage using the Weighted Least-Square (WLS) algorithm or the Multi-Dimensional Scaling-based (MDS) algorithm. Simulations and field tests show that the proposed indoor localisation scheme outperforms the existing schemes, and it can achieve an average error of 0·32 metres in an 8 m by 5 m area.

Shooter Localization based on DoA Estimation of Gunshot Signals and Digital Map Information

This paper discusses a technique to estimate the position of a sniper using a spatial microphone array placed on elevated platforms. The shooter location is obtained from the exact location of the microphone array, from topographic information of the area and from an estimated direction of arrival (DoA) of the acoustic wave related to the explosion in the gun barrel, which is known as muzzle blast. The estimation of the DOA is based on the time differences the sound wavefront arrives at each pair of microphones, employing a technique known as Generalized Cross Correlation (GCC) with phase transform. The main idea behind the localization procedure used herein is that, based on the DoA, the acoustical path of the muzzle blast (from the weapon to the microphone) can be marked as a straight line on a terrain profile obtained from an accurate digital map, allowing the estimation of the shooter location whenever the microphone array is located on an dominant position. In addition, a new approach to improve the DoA estimation from a cognitive selection of microphones is introduced. In this technique, the microphones selected must form a consistent (sum of delays equal to zero) fundamental loop. The results obtained after processing muzzle blast gunshot signals recorded in a typical scenario, show the effectiveness of the proposed method.

Development of Instrumentation for the Measurement of the Performance of Acoustic Absorbers

In both fixed and rotary wing aircraft, the move toward lighter structures has resulted in an increase in structural vibration and interior noise. Porous materials have been proposed as acoustic absorbers to reduce this noise. This paper discusses the development of equipment at the NASA Glenn Research Center for characterizing the acoustic performance of porous materials: a flow resistance apparatus to measure the pressure drop across a specimen of porous material, and a standing wave tube that uses a pair of stationary microphones to measure the normal incidence acoustic impedance of a porous material specimen. Specific attention is paid to making this equipment as flexible as possible in terms of specimen sizes need for testing to accommodate the small or irregular sizes often produced during the development phase of a new material. In addition, due to the unknown performance of newly developed material, safety features are included on the flow resistance apparatus to contain test specimens that shed particles or catastrophically fail during testing. Results of measurements on aircraft fiberglass are presented to verify the correct performance of the equipment.

Experimental characterization of vertical-axis wind turbine noise.

Vertical-axis wind turbines are wind-energy generators suitable for use in urban environments. Their associated noise thus needs to be characterized and understood. As a first step, this work investigates the relative importance of harmonic and broadband contributions via model-scale wind-tunnel experiments. Cross-spectra from a pair of flush-mounted wall microphones exhibit both components, but further analysis shows that the broadband dominates at frequencies corresponding to the audible range in full-scale operation. This observation has detrimental implications for noise-prediction reliability and hence also for acoustic design optimization.

Subband processing‐based approach for the localisation of two simultaneous speakers

Time difference of arrival (TDOA)-based techniques are a main category of speaker localisation methods. In a large subcategory of these methods, the generalised cross-correlation (GCC) is employed for TDOA estimation. In this study, the authors propose a subband processing-based method that computes the GCC of the microphone pairs in each subband. The information collected from different subbands is then combined together to estimate the direction of two simultaneous speakers. While the conventional methods consider the whole signal spectrum in the localisation procedure, the proposed method takes advantage of the difference in the frequency contents of the speakers. The proposed method computes the histograms of the peak positions of the GCC curve for each microphone pair in different subbands. These histograms are then fused using one of the three proposed histogram averaging methods, called simple, sectional, and weighted averaging. The proposed method has been evaluated on simulated and real speech data in noisy, reverberant, and noisy–reverberant conditions. The evaluation results demonstrate the superiority of the proposed subband processing-based method over its full-band counterpart. The authors’ experiments also show that among different histogram averaging methods, the weighted averaging has greater performance in estimating the direction of speakers.

English

Source localisation is a method to estimate position of a source. In case of acoustic source localisation (ASL), the location of sound source is estimated using acoustic sensors such as a microphone. In case of ASL, time difference of arrival (TDOA) from each pair of microphones is estimated. For any pair of microphones, the surface on which the TDOA is constant is a hyperboloid of two sheets. Then the source location is estimated at the point where all associated hyperboloids most nearly intersect. This concept has been used in our range in finding the point-of-burst of artillery shell using an array of sensors. In this paper, a simulation model has been developed to examine the applicability of acoustic source localisation for determining point-of-burst of artillery shells. The randomness in the model has been incorporated in terms of gustiness of downrange sea wind. The result of the simulation has been validated with trajectory data of projectiles tracked by radar. Finally, an acoustic sensor array-based setup has been developed and used for localising point-of-bursts. Science Journal, Vol. 64, No. 6, November 2014, pp.517-523, DOI:http://dx.doi.org/10.14429/dsj.64.8112

Distributed Kalman filter-based speaker tracking in microphone array networks

Using a microphone array network, a speaker tracking method based on distributed Kalman filter (DKF) in a noisy and reverberant environment is proposed. Firstly, the time delay of arrival (TDOA) in each microphone pair is estimated by the generalized cross-correlation (GCC) method. Next, the Langevin model is used as state equation to model the speaker’s movement, meanwhile the measurement equations with true TDOA are deduced by linearizing the TDOA model. Finally, the moving speaker’s positions are estimated by distributed Kalman filtering in a microphone array network. The proposed method is scalable. It can obtain a trajectory of the speaker’s movement smoothly with excellent tracking accuracy. Simulation results verify the effectiveness of the proposed method.

Joint estimation of pitch and direction of arrival: improving robustness and accuracy for multi-speaker scenarios

In many speech communication applications, robust localization and tracking of multiple speakers in noisy and reverberant environments are of major importance. Several algorithms to tackle this problem have been proposed in the last decades. In this paper, we propose several extensions to a recently presented joint direction of arrival (DOA) and pitch estimation method, increasing its robustness in multi-speaker scenarios, noise, and reverberation. First, a spectral comb filter is added to the original algorithm to better cope with concurrent speakers. Second, the well-known generalized cross-correlation with phase transform (GCC-PHAT) is used as an additional weighting function to improve the DOA estimation accuracy in terms of correct hits. Third, using multiple microphone pairs, the multi-channel cross-correlation approach is incorporated to improve the robustness against noise and reverberation. In order to improve tracking for moving and even intersecting speakers, a particle filter is used. Experiments with real-world recordings in realistic acoustic conditions show that the proposed extensions increase the DOA hit rate by about 33% compared to the original algorithm for two step-wise moving sources at a signal-to-noise ratio (SNR) of 15 dB and a reverberation time RT60 of 560 ms.

Sub-sampling-based 2D localization of an impulsive acoustic source in reverberant environments

This paper presents a robust method for two-dimensional (2D) impulsive acoustic source localization in a room environment using low sampling rates. The proposed method finds the time delay from the room impulse response (RIR) which makes it robust against room reverberations. We consider the RIR as a sparse phenomenon and apply a recently proposed sparse signal reconstruction technique called orthogonal clustering (OC) for its estimation from the sub-sampled received signal. The arrival time of the direct path signal at a pair of microphones is identified from the estimated RIR, and their difference yields the desired time delay estimate (TDE). Low sampling rates reduces the hardware and computational complexity and decreases the communication between the microphones and the centralized location. Simulation and experimental results of an actual hardware setup are presented to demonstrate the performance of the proposed technique.

Passive acoustic method for aircraft states estimation based on the Doppler effect

A passive acoustic method to locate moving sound sources is applied to maneuvering aircraft. The advantages of the method are that it is suitable for all kinds of aircraft, not only propeller-driven, and is not restricted to low height above the ground. Its applicability could be, for instance, to supplement aircraft noise monitoring systems or to supervise small airports' activities. The method is based on the relation between the relative Doppler effect observed from a set of at least seven microphone receivers, distributed in the airport surroundings, and the aircraft position and speed. The method requires knowledge of the position of the aircraft at the start of takeoff. The ambiguity function is used to calculate the relative Doppler stretch of the spectrum of the sound between pairs of microphones. The results of applying the ambiguity function are the inputs to a system of equations that relates the aircraft position and speed to the relative Doppler frequency stretches. This system of equations is solved using a genetic algorithm. The performance of the method was tested by computer simulation. The results showed that the location errors are of the same order of magnitude as the size of an average aircraft, even if the takeoff position is not accurately known.

Mosquito (Aedes aegypti) flight tones: frequency, harmonicity, spherical spreading, and phase relationships.

Mosquito flight produces a tone as a side effect of wing movement; this tone is also a communication signal that is frequency-modulated during courtship. Recordings of tones produced by tethered flying male and female Aedes aegypti were undertaken using pairs of pressure-gradient microphones above and below, ahead and behind, and to the left and right over a range of distances. Fundamental frequencies were close to those previously reported, although amplitudes were lower. The male fundamental frequency was higher than that of the female and males modulated it over a wider range. Analysis of harmonics shows that the first six partials were nearly always within 1 Hz of integer multiples of the fundamental, even when the fundamental was being modulated. Along the front-back axis, amplitude attenuated as a function of distance raised to the power 2.3. Front and back recordings were out of phase, as were above and below, while left and right were in phase. Recordings from ahead and behind showed quadratic phase coupling, while others did not. Finally, two methods are presented for separating simultaneous flight tones in a single recording and enhancing their frequency resolution. Implications for mosquito behavior are discussed.

Passive acoustic localization of small aircraft

Stevens Institute of Technology has built the Acoustic Aircraft Detection (AAD) system for the detection, tracking and classification of Low Flying Aircraft (LFA). LFA may be of concern as they have been used for illicit operations. The AAD consists of several nodes deployed in a wide area, where each node acquires signals from a cluster of five microphones. The calculation of the cross-correlation function of acoustic signals received by various microphone pairs is applied for finding the direction of the signal arrival. Fusion of time difference of arrival estimates from several pairs of acoustic sensors has resulted in the improvement of angle estimation accuracy. Triangulation of the direction of arrival estimates from two or more nodes was applied for determining the target position and altitude. Kalman filtering was used for smoothing the target tracks and decreasing the localization uncertainties. Software for predicting the performance was developed and was used to inform sensor placement in the field tests. The field tests were conducted with various kinds of LFA—single-engine, helicopters, and ultralights. Comparison of the acoustic tracking results with the GPS ground truth showed that the errors are similar to the theoretical predictions. [This work was funded by DHS S&T.]

Infrasonic emissions from aircraft wake vortices: Experimental results

Infrasonic emissions from aircraft wake vortices were investigated at the Newport News-Williamsburg International Airport early in the year 2013. Signals received by the microphones situated along an airport runway were processed in 10-s intervals. As an aircraft accelerates toward takeoff, it produces a large pressure burst as it passes each microphone. Following the burst, there appear low-frequency signals of high coherence among microphone pairs. These are interpreted as emissions from the aircraft wake vortices, as suggested by theory. In successive 10-s intervals, the coherence gradually diminishes to background levels, signifying the disappearance of the vortices. On landing the intervals of high coherence precede the bursts at aircraft touchdown, and then diminish. The pressure burst serves as a time stamp for the ensuing vortex emissions and thereby permits the tracking of successive takeoff or landing events on the same runway or on adjacent runways. The emission spectrum is essentially broadband, lacking spectral features (e.g., tones). Data were taken for takeoff of Airbus 319, DC-9, MD-88, CRJ, Lear Jet, Corporate Jet, and Dash-8 aircraft, and for landing of the Airbus 319. The pattern of pressure burst, high-coherence intervals, and diminishing-coherence intervals was observed for all takeoff and landing events without exception.

Algorithms for Acoustic Source Orientation Based on Microphone Array in Robot Audition

Acoustic source orientation is an important feature in robot audition. This paper applied a spatial cone six-element (SCSE) microphone array and combined with the time difference of arrival (TDOA) between pairs of spatial separated microphones to estimate acoustic source orientation. Simulate three stages of the acoustic orientation process in a real indoor environment, and results show that the algorithm is simple and effective, reducing the amount of calculation, having anti-noise and reverberation ability to meet the requirements of orientation accuracy.

Investigation of noise radiation from a swirl stabilized diffusion flame with an array of microphones

Next generation of combustors are expected to provide significant improvement on efficiency and reduced pollutants emission. In such combustors, the challenges of local flow, pressure, chemical composition and thermal signatures as well as their interactions will require detailed investigation for seeking optimum performance. Sensor networks with a large number of sensors will be employed in future smart combustors, which will allow one to obtain fast and comprehensive information on the various ongoing processes within the system. In this paper sensor networks with specific focus on an array of homogeneous microphones are used examine the spectral characteristics of combustion noise from a non-premixed combustor. A non-premixed double concentric swirl-flame burner was used. Noise spectra were determined experimentally for the non-premixed swirl flame at various fuel–air ratios using an array of homogeneous condenser microphones. Multiple microphones positioned at discrete locations around the turbulent diffusion flame, provided an understanding of the total sound power and their spectral characteristics. The growth and decay coefficients of total sound power were investigated at different test conditions. The signal coherence between different microphone pairs was also carried out to determine the acoustic behavior of a swirl stabilized turbulent diffusion flame. The localization of acoustic sources from the multiple microphones was examined using the noise spectra. The results revealed that integration of multiple sensors in combustors and reactors does allow one to obtain additional information which can be further tailored for the quest to develop advanced combustors.

Localization of Acoustic Reflective Boundary Using a Pair of Microphones and an Arbitrary Sound Source

Compassing the spatial shape is sometimes performed to develop a surround speaker system or enhance a signal. Although the spatial shape is reconstituted from the boundary position, conventional methods offer arrays consisting of a large number of elements (microphones and loudspeakers). In this research, we developed a method of localizing the acoustic reflective boundary based on the direction-of-arrival (DOA) of direct and reflected sounds using a pair of microphones. This method offers the DOA values of direct and reflected arbitrary sound sources. Although the estimation involves some estimation errors caused by approximations in the calculation, the correction function was obtained from several conditions numerically. Results indicate that it is possible to localize the reflective boundary almost accurately (the average error ratio from the true value is below 10%) through the presented method.

Position-Dependent Crosstalk Cancellation Using Space Partitioning

The present study tested a new stereo playback system that effectively cancels cross-talk signals at an arbitrary listening position. Such a playback system was implemented by integrating listener position tracking techniques and crosstalk cancellation techniques. The entire listening space was partitioned into a number of non-overlapped cells and a crosstalk cancellation filter was assigned to each cell. The listening space partitions and the corresponding crosstalk cancellation filters were constructed by maximizing the average channel separation ratio (CSR). Since the proposed method employed cell-based crosstalk cancellation, estimation of the exact position of the listener was not necessary. Instead, it was only necessary to determine the cell in which the listener was located. This was achieved by simply employing an artificial neural network (ANN) where the time delay to each pair of microphones was used as the ANN input and the ANN output corresponded to the index of cells. The experimental results showed that more than 95% of the experimental listening space had a CSR ≥ 10 dB when the number of clusters exceeded 12. Under these conditions, the correlation between the true directions of the virtual sound sources and the directions recognized by the subjects was greater than 0.9.

On the statistical distributions of active intensity directions

Pressure and pressure differences measured using closely-spaced microphone pairs enable the calculation of active intensity directions. These directions are exploited in various applications such as acoustical analysis and modelling, multichannel audio systems, source localization and real-time blind source separation. Despite their widespread usage, the distribution of the active intensity directions in the presence of interference has not been investigated before, except for the diffuse sound field case for which the distribution becomes uniform. In this paper, the interference problem has been formulated using the pressure and pressure differences. Algebra of random variables, fractional linear transformations and Brownian motion have been used for the exposition of the model distributions. It is shown that the distribution is wrapped Cauchy for a single sound source in the presence of independent interference in an anechoic chamber. In a direct plus diffuse sound field, the distribution is offset Cauchy for a nonstationary target source and von Mises for the special case of a stationary target source, related to the Brownian motion. In the presence of multiple independent sound sources, the distribution can be modeled as mixtures of circular Cauchy distributions and a uniform distribution.