Estimation of the fractal parameter from noise data using wavelet
Shou Zhang · Jounal of Xidian University · 2000
The estimation of Hurst index from noisy fractal signals is an important problem. In this paper, a novel Hurst index estimation method based on the DFGN model and wavelet analysis is proposed. We have derived the formulas for wavelet coefficients of the fractal signal and white noise separately on the basis of which the maximum likelihood method is used to estimate the fractal parameter. The simulation experiments show that the method is more reliable than the traditional method.