Scholarlink

<p>An efficient signal classification algorithm for detecting Gaussian noise by the values of autocorrelation and wavelet autocoherence coefficients is presented. A comparative analysis of the autocorrelation methods for time series analysis and the wavelet autocoherence method applied to time series of large-scale wavelet coefficients is carried out. A database of 20 types of model signals (both linear and nonlinear frequency modulations) is used, which significantly expands the possibilities of using the algorithm in automatic data recognition systems. According to the results of the study, the value of the autocoherence coefficient remains unchanged over the entire range of noise power changes, while the value of autocorrelation depends on the frequency modulation and is different. To obtain a simplified model, the Shapiro–Wilk test (W-test) was used, and signals are classified into two distinct groups according to the values of the autocorrelation and wavelet autocoherence coefficients. The noise threshold is determined for the signals that correspond to the normal law of data distribution.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Analysis of Time Series Using Wavelet Autocoherence and Autocorrelation

  • 期刊论文

  • 发表日期：

    2025年3月1日

查看全文
  • Yu. Taranenko,
  • O. Oliinyk,
  • B. I. Moroz,
  • V. Lopatin

摘要

An efficient signal classification algorithm for detecting Gaussian noise by the values of autocorrelation and wavelet autocoherence coefficients is presented. A comparative analysis of the autocorrelation methods for time series analysis and the wavelet autocoherence method applied to time series of large-scale wavelet coefficients is carried out. A database of 20 types of model signals (both linear and nonlinear frequency modulations) is used, which significantly expands the possibilities of using the algorithm in automatic data recognition systems. According to the results of the study, the value of the autocoherence coefficient remains unchanged over the entire range of noise power changes, while the value of autocorrelation depends on the frequency modulation and is different. To obtain a simplified model, the Shapiro–Wilk test (W-test) was used, and signals are classified into two distinct groups according to the values of the autocorrelation and wavelet autocoherence coefficients. The noise threshold is determined for the signals that correspond to the normal law of data distribution.