Aliased Frequency Estimation by Using Two Coprime Sub-Nyquist Sampling Rates

In this brief, we propose a sub-Nyquist frequency estimation method, which aims to increase the instantaneous bandwidth of discrete Fourier transform (DFT) based digital wideband receiver and estimate the frequencies without ambiguity. The proposed method consists of two parallel coprime sub-Nyquist sampling channels. We firstly address the frequency estimation based on the outputs of DFT, and the no-overlapping frequency components (FCs) can be estimated without grid. To obtain gridless estimation, the overlapping FCs are then reformulated as an finite rate of innovation problem with lower rate of innovation, larger sampling interval and higher signal-to-noise rate after removing the no-overlapping FCs via partial channel reconstruction, and are estimated with smaller errors by the sparse samples. Numerical simulations demonstrate that the proposed method is effective and noise robust.