Direction of Arrival Estimation for Mixed Near-Field and Far-Field Targets Based on Chirplet Transform

When estimating the direction of arrival (DOA) of targets with a linear array, directly applying DOA estimation models based on the far-field (FF) assumption to near-field (NF) targets can cause mismatches in the array processing model. This leads to DOA estimation errors, lower array gain, and missed NF targets. To address these issues, we propose a method based on chirplet transform for joint DOA estimation of mixed FF and NF targets with a linear array. The basic idea is that each element of the array manifold vector can be treated as linear frequency modulated signals. NF signals correspond to array manifold vectors with nonzero frequency modulation slopes, while FF target signals correspond to array manifold vectors with zero frequency modulation slopes. By introducing time-frequency analysis methods, chirplet transform is used to estimate the azimuth of unknown targets. In comparison with high-order mixed-field target DOA estimation methods (two-stage MUSIC, oblique projection MUSIC, and spatial differencing algorithm), the proposed method is not restricted by element spacing, does not suffer from array aperture loss, and can be used for coherent signal DOA estimation. In addition, compared with conventional beamforming (CBF) and focused beamforming, the proposed method provides higher spatial gain and better target resolution, respectively. Finally, numerical simulations and lake trial data under the same test conditions show that the proposed method improves the output signal-to-noise ratio of NF target signals by more than 8 dB compared with CBF.