New Algorithms for Adaptive Notch Smoothing

The problem of extraction/elimination of a nonstationary complex sinusoidal signal buried in noise is considered. This problem is usually solved using adaptive notch filtering (ANF) algorithms. It is shown that accuracy of signal estimation can be increased if the results obtained from ANF are further processed using a cascade of appropriately designed filters. The resulting adaptive notch smoothing (ANS) algorithms can be employed in many offline or nearly real-time online applications. Whenever signal frequency varies in a sufficiently smooth manner, the proposed fixed-interval and fixed-lag ANS algorithms, based on a new, quasi-linear model of frequency changes, outperform the existing solutions.