A software tool for compressive sensing based time-frequency analysis

摘要

A software tool that implements Compressive sensing based time-frequency analysis and performs instantaneous frequency estimation, is proposed and described in the paper. A focus is made on the signals with fast varying instantaneous frequency (IF), which can be accurately estimated using complex-time distribution. Therefore, the proposed tool offers different possibilities to adjust parameters of complex-lag distributions in order to comply with fast-varying IF laws. Moreover, beside the standard implementation based on the full set of samples, a compressive sensing based time-frequency approach is included in order to obtain sparse time-frequency representation. Sparse time-frequency representation is reconstructed from very few ambiguity domain observations. The tool performance is tested on real and synthetic signals.