Signal estimation method based on first-order differential microphone array

Abstract

The invention provides a signal estimation method based on a first-order difference microphone array, and the method comprises the steps: firstly constructing a first-order difference microphone array model, receiving signals through the array, and carrying out the subtraction of the signals received by two microphones, thereby obtaining a time domain difference signal; and then frequency domain analysis is carried out on the difference value signal, it can be found that an amplitude-frequency characteristic curve of the difference value signal has a comb-like structure, and the effect is called a comb filtering effect in the invention. And mathematical analysis is carried out on the effect, and a reverse comb filter is designed according to the mathematical analysis. And filtering the difference signal by using the filter to restore the frequency spectrum of the original signal, and finally restoring the original time domain signal in combination with the phase spectrum to achieve the purpose of signal estimation. The method can effectively solve the problems that a traditional signal estimation method is low in accuracy, and a novel signal estimation method based on deep learning is high in calculation complexity and poor in robustness.

Description

technical field [0001] The invention belongs to the technical field of signal processing, and in particular relates to a signal estimation method based on a first-order differential microphone array. Background technique [0002] Signal estimation is the first step of the signal processing system, and the quality of the signal obtained by the estimation will directly affect the performance of the processing system. At present, the technical research on signal estimation can be roughly divided into two categories: [0003] One is the traditional signal spectrum estimation method based on the Fourier transform of the correlation function and other technologies. The performance of this method depends on the known data sequence length, the frequency resolution is low, and it is not suitable for short-term data. It is suitable for estimating stationary signals, and most of the signals in the real environment are non-stationary signals. In these cases, the estimation accuracy of ...

AI Technical Summary

Problems solved by technology

[0006] In view of this, the present invention aims to provide a signal estimation method based on a first-order difference microphone array. Since the method is designed based on a physical propagation model, it can effectively solve the problem of low accuracy of traditional signal estimation methods and problems based on The problem of high computational complexity and poor robustness of new signal estimation methods for deep learning

Images