Error surface analysis of adaptive comb filter

A minimal parameter infinite-impulse-response (IIR) adaptive comb filter model is presented for use in eliminating or retrieving a harmonic series buried in a broadband signal. The adaptive comb filter estimates and tracks the fundamental frequency of the harmonic series. Error surface analysis is carried out to obtain the convergence characteristics of the proposed comb filter. Special pole and zero constraints are introduced which enhance the filter performance. Results show that although the error surface is multimodal its performance is excellent for a certain class of filtering applications.