Very often, when a signal exhibits some periodicity, it is modeled by a single or a combination of sinusoids buried in noise. In such a modeling, signal characteristics including amplitudes, phases and frequencies may also vary with time. Estimation of the frequencies and extraction of the individual sinusoidal components of such a signal provide useful information about the signal, and therefore provide means for signal analysis. Introducing algorithms capable of decomposing such a signal into its individual frequency components is an active subject of research and finds applications in various branches of engineering. In this thesis, a modified adaptive notch filter (ANF) scheme is proposed that directly estimates the frequency of sinusoidal signals. The ANF is a second-order notch filter that is further furnished with a nonlinear differential equation to update the frequency. This scheme is then extended to a structure composed of coupled paralleled ANFs that estimates directly the frequencies of a signal compos ......
