Skip to Main Content
In this paper, some nonstationary properties, e.g., self-similarity, long-term dependency of wavelet transform coefficients of fractal signal and noise at different decomposition scales are analyzed. Based on the minimum mean square error of these wavelet coefficients at each scale, a new method of estimating fractal signal from additive white noise is proposed in pervasive computing environment. The parameters of the background noise in this method can be dynamically adapted in runtime to model the variation of both the signal and the noise. Since it doesn't need to know the parameters of fractal signal and the statistical characteristic of added white noise in advance, this method is suitable in various situations. The simulation results show that this method has good performance to be used in pervasive computing environment.