Adaptive filter structures for deconvolution of seismic signals

Abstract

The results of application of three different methods to the "adaptive deconvolution" of seismic data are reported here. These are based on the adaptive tapped delay line filter, the adaptive lattice filter and the adaptive Kalman-filter identifier. All the three methods are shown to be superior to the "fixed-structure" predictor. A comparison of the three methods shows that the adaptive lattice structure has faster convergence than the adaptive tapped delay line structure. The adaptive Kalman filter identifier is found to be better than both the others, although it is computationally more complex. The conclusions are based both on theoretical studies as well as on experiments with real seismic data.