Adaptive digital filters

Both analog and equivalent digital filters having adaptive capability are discussed. The analysis is restricted to second order sections but is extendable. Simple to implement gradient procedures using energy or phase information are used. Neither require supervised adaption. Applications include tracking nonstationary signals such as in seismic and acoustical sensing. Aside from the results shown, the tutorial nature of the paper establishes the need for further study to determine the optimum transformation that will minimize the interaction of the analog parameters (center frequency and bandwidth) in the transformed digital domain. It is suggested that this be realized by adaptive decorrelation, modification of the topological structures, and design of sm appropriate equivalent transformation.