透過您的圖書館登入
IP:3.138.125.2

摘要


A technique of cross-term suppression in WD (Wigner Distribution) for a multi-component signal that is embedded WGN (White Gaussian Noise) is proposed. In this technique, an optimized algorithm is developed for time-varying noisy signal and a CAD (Computer Aided Design) simulator is designed for Numerical simulations of synthetic signal. In proposed technique, signal components are localized in t-f (time frequency) plane by STFT (Short Time Fourier Transform). Rectified STFT is computed and Spectral Kurtosis is used to separate a signal components from noise in t-f plane. The t-f plane is segmented and then signal components are filtered out by FFT (Fractional Fourier Transform). Finally, WD (free of cross terms) of isolated signal component is computed to obtain high resolution in t-f plane.

延伸閱讀