Abstract
Real-time flow velocity measurement is a practical issue in industrial and biomedical applications. Because their good frequency resolution, parametric methods such as autoregressive (AR) modeling and time-frequency distributions (TFD) are generally preferred to Fourier analysis. However, these methods become highly inaccurate in the presence of colored noise. We review here the principal parametric and nonparametric techniques and show their limitations in the estimation of Doppler frequency in the presence of strong colored noise. Different solutions to overcome these limitations are then proposed and compared using synthetic Doppler signals with colored noise.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
