Nonparametric spectral analysis of gapped data via an adaptive filtering approach (Q5953484)

The paper proposes an algorithm for complex spectral analysis of a data sequence with missing samples (or gaps). Such situations can be found for example in astronomical applications or in radar signal processing. The algorithm is based on a nonparametric adaptive finite impulse response (FIR) filtering spectral estimation approach and consists in a cyclic procedure alternating a spectrum estimation step (using the most recent estimate of the missing data samples) and a gap filling step (using the most recent spectrum estimate).