A Chirp-based, Adaptive, Signal-dependent Reduced Interference Distribution for Limited Data

Noise-like artifacts, which are caused by incomplete and randomly sampled data, spread over the whole ambiguity domain, and thus seriously obscure the true time-frequency signature of the data. In this paper, a new design for the signal-dependent adaptive kernel is proposed, which is robust with missing data. The method relies on the properties of chirps whose auto-terms only reside in a fixed half of the ambiguity domain. The important thing is that this half excludes the Doppler axis, where the chirps’ noise-like artifacts concentrate. By cutting out this region when performing the optimization problem, a better signal-dependent kernel for chirps is obtained, which efficiently suppresses not only the cross-terms but also the missing sample artifacts. Moreover, since any windowed non-stationary signals can be approximated as a sum of chirps, the proposed approach can be applied to other types of non-stationary signals. It is shown in the simulation that our method outperforms other reduced interference time-frequency distributions of incomplete observations.