Alleviating Border Effects in Wavelet Transforms for Nonlinear Time-varying Signal Analysis

Advances in Electrical and Computer Engineering

SU, H. ; LIU, Q. ; LI, J..

关键词: convolution;    Fourier series;    frequency estimation;    spectrogram;    wavelet transforms;   

DOI  :  10.4316/AECE.2011.03009

学科分类：计算机科学（综合）

来源: Universitatea "Stefan cel Mare" din Suceava

【 摘 要 】

Border effects are very common in many finite signals analysis and processing approaches using convolution operation. Alleviating the border effects that can occur in the processing of finite-length signals using wavelet transform is considered in this paper. Traditional methods for alleviating the border effects are suitable to compression or coding applications. We propose an algorithm based on Fourier series which is proved to be appropriate to the application of time-frequency analysis of nonlinear signals. Fourier series extension method preserves the time-varying characteristics of the signals. A modified signal duration expression for measuring the extent of border effects region is presented. The proposed algorithm is confirmed to be efficient to alleviate the border effects in comparison to the current methods through the numerical examples.

【 授权许可】

Unknown