Time-frequency representation based chirp-like signal analysis using multiple level crossings

The instantaneous frequency of a chirp-like signal varies with time. The sampling density of a signal should preferably correspond to its local bandwidth. The paper discusses the level-crossing sampling principle, which provides the capability for performing such an analog-to-digital conversion. The properties of level-crossing sampling are considered. As the captured samples are spaced non-uniformly, appropriate digital signal processing is required. The chirp-like signal is characterized by time-frequency representation. Short-time Fourier transform and Wigner-Ville distribution are inspected and enhanced to make them applicable to signal processing in the level-crossing sampling case. Additional benefits are obtained if the combined approach is used - artifacts are removed and localization of the chirp-like signal is improved. The simulation results are demonstrated.