A channelization-based multi-band sampling method in the fractional Fourier domain for frequency estimation

Abstract

In order to reduce the difficulty of synchronizing multi-coset sampling in hardware implementation, expand the range of unambiguous frequency estimation and reduce the frequency estimation error, a channelization-based multi-band (CBMB) sampling method in the fractional Fourier domain for frequency estimation is proposed. This method retains the unambiguous phase information of multi-coset sampling while reducing implementation difficulty by increasing the sampling interval between channels and employing analog modulation. To achieve unambiguous sampling with the CBMB architecture, we analyze the fractional Fourier transform of the undersampled signal after channelized filtering and provide a method for setting undersampling parameters in the fractional Fourier domain. A sparse reconstruction and frequency estimation method based on the signal and covariance matrix is derived. Simulation analysis verifies the feasibility and effectiveness of this method, compared to existing methods, it decreases the demand for analog-to-digital converter undistorted quantization bandwidth during sampling, reduces the frequency estimation error and improves the frequency estimation accuracy of linear frequency modulated signals under the same signal-to-noise ratio.