Joint Estimation of Single Target’s High Amplitude Difference Doppler Frequencies in FMCW Radar

This article deals with the complex task of estimating Doppler frequencies in Frequency-Modulated Continuous Wave (FMCW) radar that uses single-input and multiple-output antennas. It focuses on scenarios where multiple targets are present at different locations, and each target introduces two distinct Doppler shifts with substantial difference in their signal amplitudes. The fast-Fourier transform (FFT) cannot exploit the complete data available for such a problem, resulting in poor angular and Doppler resolution along with significant side-lobe levels. In contrast, the subspace-based algorithm MUltiple SIgnal Classification (MUSIC) and Yule-Walker autoregressive (Yule-AR) methods can exploit the entire data and provide good angular and Doppler resolution with lower side-lobes, but their performance deteriorates at low signal-to-noise ratios and when the targets are close. To address these challenges, we propose a novel algorithm where the received samples are rearranged into a form that enables the application of high-resolution Capon and APES (amplitude and phase estimation)-like algorithms for angular location estimation. Next, both frequency terms are coupled, independent of the amplitude of individual frequency components, to estimate them jointly. Simulation results show that the proposed algorithm outperforms the FFT, MUSIC, and Yule-AR algorithms in terms of Doppler resolution and side-lobe level reduction.