Conference_programme: 21: Source identification and localization

Lecture: A dual-channel correlated noise reduction algorithm based on the coherence function and the continuous wavelet transform

Author(s): Di Marco Alessandro, Camussi Roberto

Summary:
A dual-channel noise reduction method is proposed in this paper and the performances evaluated with experimental data. It is based on the coherence function and the continuous wavelet transform. The algorithm is applied to the cross-wavelet coefficients obtained from the input signals. The adopted method can treat noisy signals using two sensors opportunely spaced in different noise scenarios. For the present study, fluctuating pressure signals, acquired from two microphones, are used. \nCoherence-based enhancement methods require a reference signal to be correlated with the noisy signal. The performance of the method depends on the characteristics of the environmental noise. In this case, the post-processing method aims at decontaminating the pressure time signal from any coherent noise source. This situation may occur in wind tunnel testing when the transducers used for the acquisition are simultaneously subjected to the same well-defined background noise. \nThe analysed data are obtained from measurement campaigns opportunely designed to reproduce an analogous test case and/or previously conducted by the authors.\nThe pressure signals from the experiments are decontaminated using three algorithms: a standard method in frequency domain, the coherent output spectrum technique, and two algorithms in the time-frequency domain. A series of tests have been conducted to examine the efficacy and limitations of each of the procedures for specific applications. The results are compared to show the main practical advantages. It is found that he proposed algorithm is robust and does not require large computational complexity; furthermore, the output is a clean signal in the physical domain rather than a clean spectrum. The latter aspect is of particular importance because it allows the user to apply further methods, like beamforming, to obtain a higher signal-to-noise ratio.\n

Corresponding author

Name: Dr Alessandro Di Marco

e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.

Country: Italy