High-precision frequency estimation for FMCW radar applications based on parameterized de-alternating and modified ICCD

Abstract

In this paper, aiming to achieve high-resolution frequency estimation for frequency-modulated continuous wave (FMCW) radar applications, a novel frequency estimation algorithm based on parameterized de-alternating (PDA) is proposed. According to the beat signal model, the proposed PDA method is performed with a matched kernel function to achieve the best de-alternating performance (i.e. the maximum conversion from an oscillating signal to a constant signal). The kernel parameter is estimated by solving an optimization problem which is simple and computationally efficient. Moreover, for a multiple target scenario, the intrinsic chirp component decomposition (ICCD) method is employed to isolate different components of the beat signal, which can eliminate the mutual interference between adjacent components when performing the PDA method. Compared with the commonly used FFT technique, our method can greatly improve the frequency estimation resolution and achieve superior ranging performance with a limited bandwidth. Both simulated and experimental validations are provided to show the effectiveness of the proposed methods.