Detection of weak astronomical signals with frequency-hopping interference suppression

This paper addresses the detection of weak astronomical signals that are contaminated by strong frequency-hopping (FH) interferers and suffer from missing samples. The problem is considered in the time-frequency domain and we successively suppress artifacts due to missing samples, estimate and remove FH interferers, and detect the weak astronomical signals. More specifically, we first suppress the artifacts due to missing samples by developing a waveform-adaptive time-frequency kernel. The instantaneous spectra of the FH interferers are then estimated using a sparsity-based approach that takes the inherent properties of FH signals into account. Finally, a sparse coherent integrated cubic phase function is applied to effectively detect weak astronomical chirp components over a long integration time. Simulation results are provided to demonstrate the effectiveness of the proposed approach.