<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>The echo signals of the radar in complex backgrounds are often very unstable and thus require effective noise cancellation. In this paper, according to the characteristics of continuous wavelet variation and discrete wavelet variation, the decomposition effect of multi-resolution analysis and orthogonal Mallat algorithm on low-frequency and high-frequency non-smooth signals is studied, and the selection method of wavelet bases is explored. Then, the noise characteristics affecting the pulsed LIDAR system are analyzed, and the LIDAR pulse signal is simulated by MATLAB, while Gaussian white noise is introduced to obtain the noise-added echo signal, and then multiple wavelet threshold denoising methods are applied to denoise the echo signal. For the input signal-to-noise ratio of −10.57 dB, the output signal-to-noise ratios of db8, db9, db10, and bior3.5 wavelet bases under forced thresholding are −1.971, −2.178, −2.173, and −1.032, respectively. For different input signal-to-noise ratios, the average root mean square error of db8, db9, db10, and bior3.5 wavelet bases under default thresholding is 1.51. The denoising methods for radar signals using the properties of wavelet decomposition have obvious superiority compared to traditional filters, and the wavelet transforms threshold denoising methods have wide adaptability.</p>
</abstract>

The echo signals of the radar in complex backgrounds are often very unstable and thus require effective noise cancellation. In this paper, according to the characteristics of continuous wavelet variation and discrete wavelet variation, the decomposition effect of multi-resolution analysis and orthogonal Mallat algorithm on low-frequency and high-frequency non-smooth signals is studied, and the selection method of wavelet bases is explored. Then, the noise characteristics affecting the pulsed LIDAR system are analyzed, and the LIDAR pulse signal is simulated by MATLAB, while Gaussian white noise is introduced to obtain the noise-added echo signal, and then multiple wavelet threshold denoising methods are applied to denoise the echo signal. For the input signal-to-noise ratio of −10.57 dB, the output signal-to-noise ratios of db8, db9, db10, and bior3.5 wavelet bases under forced thresholding are −1.971, −2.178, −2.173, and −1.032, respectively. For different input signal-to-noise ratios, the average root mean square error of db8, db9, db10, and bior3.5 wavelet bases under default thresholding is 1.51. The denoising methods for radar signals using the properties of wavelet decomposition have obvious superiority compared to traditional filters, and the wavelet transforms threshold denoising methods have wide adaptability.

Denoising complex background radar signals based on wavelet decomposition thresholding

University of Science and Technology of China

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

The echo signals of the radar in complex backgrounds are often very unstable and thus require effective noise cancellation. In this paper, according to the...

Denoising complex background radar signals based on wavelet decomposition thresholding

Online veröffentlicht: 09. Okt. 2023

Seitenbereich: -

Eingereicht: 05. Jan. 2023

Akzeptiert: 19. Apr. 2023

DOI: https://doi.org/10.2478/amns.2023.2.00535

Schlüsselwörter
Wavelet decomposition, Threshold denoising, Radar echo signal, Signal-to-noise ratio, Gaussian white noise

© 2023 Feng Qiu et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Denoising complex background radar signals based on wavelet decomposition thresholding

Online veröffentlicht: 09. Okt. 2023

Seitenbereich: -

Eingereicht: 05. Jan. 2023

Akzeptiert: 19. Apr. 2023

DOI: https://doi.org/10.2478/amns.2023.2.00535

SchlüsselwörterWavelet decomposition, Threshold denoising, Radar echo signal, Signal-to-noise ratio, Gaussian white noise

© 2023 Feng Qiu et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Schlüsselwörter
Wavelet decomposition, Threshold denoising, Radar echo signal, Signal-to-noise ratio, Gaussian white noise