Analysis of Signal Distortion Remodeling in Loud Environments

Abstract

Typically, disturbances during conversation will influence speech signals. In digital hearing aids, a Wiener filter is developed to reduce the noise signal that is mixed with the speech stream. By predicting the relationship between the power spectra of the speech signal that is influenced by noise and the noise signal, the Weiner filter plays a significant role in noise suppression and augmentation. The main issues with integrating Weiner filters into large communication systems are power consumption and hardware requirements. In this work, a fast Fourier transform (FFT) and inverse FFT procedure was used in conjunction with an effective Wiener filter to suppress noise. The iteration issues in the traditional Wiener filter were addressed by the proposed Wiener filter. In the suggested design, an effective inverse and multiplication operation took the place of the division operation. By modifying the suggested approach for speech signal noise degradation, an effective reduction in power and area was attained. Additionally, we looked into uniform quantization, quantization based on the A-Law and Mu-Law transformations, and quantization following noise addition. To eliminate noise, we used a variety of filtering methods.