Modulation Signal Parameter Estimation and Measurement and Control Technology Based on Time-varying Noise Suppression
Ping Wu, Xiao Kou, Wanyi Zhang, Jun Yang · 2024
Noise suppression is an important issue in digital communication systems. Traditional methods such as power spectral density estimation and frequency domain analysis can effectively suppress Gaussian white noise, but these methods may no longer be effective in the case of time-varying noise. This article studied methods of time-varying noise suppression, with the aim of improving the noise resistance of signal parameter estimation. This article analyzed the signal characteristics, noise characteristics, and the influence of interference sources. The theoretical basis and feasibility of parameter estimation methods based on time-varying noise suppression were determined through theoretical analysis. Through experiments, the performance of methods based on time-varying noise suppression was validated under different noise and interference conditions. The difference in the effectiveness of time-varying noise suppression between methods based on time-varying noise suppression and traditional methods (such as power spectrum estimation, autocorrelation function, frequency domain analysis, etc.) was compared to evaluate their parameter estimation accuracy and noise resistance. The experimental results of this article showed that when the signal-to-noise ratio was 50 decibels (dBs), the power spectral density estimation denoising rate (49.84%), autocorrelation function denoising rate (56.06%), frequency domain analysis denoising rate (66.85%), and time-varying noise suppression denoising rate (92.34%) were compared. The time-varying noise suppression method had the highest denoising rate, indicating that its denoising effect was the best. The results of this article contribute to improving the noise and disturbance resistance of measurement and control systems in complex environments, enhancing system stability and reliability, and promoting their engineering applications.