Time-Frequency Analysis of Geoacoustic Data Using Adaptive Matching Pursuit

The paper presents the results of applying a developed approach based on sparse approximation methods for time-frequency analysis of geoacoustic data. A sparse representation of the geoacoustic signal and the developed adaptive matching pursuit algorithm are described using the combined dictionary of Berlage and Gauss functions. The results of the time-frequency analysis of geoacoustic signals are presented. The features revealed by the proposed approach that arise in the time-frequency structures of geoacoustic data before earthquakes are shown. The article was prepared based on materials in a report at the Third All-Russian Acoustics Conference (September 21-25, 2020, St. Petersburg).