Comparison of microphone array denoising techniques and application to flight test measurements

Abstract

This paper deals with the denoising of microphone array measurements. In many situations, flush mounted microphone arrays are polluted by a turbulent boundary layer, this is typically the case considering wind tunnels or inflight tests. Acoustic imaging results are strongly affected by this noise, classical approaches to solve this issue consist in removing the diagonal terms from the measured cross spectral matrix, or to implement background noise subtraction strategies. This can be sufficient for conventional beamforming approaches, but can be a limitation when implementing more advanced identification methods. This work introduces two alternative techniques, a first one based on a statistical model whose parameters are inferred from measurements (PFA-Probabilistic Factor Analysis), and a second one based on the use of noise-free references. The former is an original contribution of the work, while the later is a well known approach yet not often used in the present context. Both methods, as well as more classical approaches, are compared in the frame of inflight array measurements for the characterization of jet noise. It is shown that the proposed advanced denoising approaches show enhanced performances as compared to classical approaches when applying either conventional beamforming or inverse source characterization.

A probabilistic approach for cross-spectral matrix denoising: Benchmarking with some recent methods

“…Finally, the Bayesian-based method PFA allows for a high level of denoising, assessed on numerical simulations and on measurements in presence of a strong TBL noise, as shown in the present study and in Ref. 48. The main drawback of the PFA method is its high computational cost, yet this is balanced by some benefits, among which:…”

Performance Enhancement of Functional Delay and Sum Beamforming for Spherical Microphone Arrays

“…(4) Arrange the focus points on a spherical surface 1 m away from the array center with the spacing ∆θ = 5 • and ∆ϕ = 5 • , respectively, thus there creating a total of 37 × 72 focus points. (5) Calculate the EFDAS output of each focus point according to Equation (14), and thus perform the acoustic imaging.…”

“…In recent years, the denoising methods based on the reconstructed CSM have been proposed, which can suppress noise interference while maintaining the completeness of the CSM. The CSM reconstruction methods, such as diagonal reconstruction (DRec), robust principal component analysis (RPCA), and probabilistic factor analysis (PFA), have been extensively studied in planar microphone arrays [14,15], but their applications in spherical arrays have not been reported. DRec, reported by Dougherty [16], minimizes the sum of the diagonal elements of the measured CSM under the constraint that the denoised CSM is positive semidefinite, and is solved by linear programming.…”

A Comparison Between Different Wall Pressure Measurement Devices for the Separation and Analysis of TBL and Acoustic Contributions

“…Previous applications of PFA for denoising have shown promising results in the medium frequency range [12,13,24], with a lack of performance in the low frequency domain, where the TBL noise is highly correlated over the microphone. This problem is here overcome by taking into account a TBL noise contribution in the inference problem.…”