In this paper, we propose a fast and robust parameter estimate method for multi-component linear frequency modulated (LFM) signals from a finite number of noisy discrete time observations. First, a new kernel called generalized adjustable parameter correlation kernel (GAPCK) is introduced to avoid the coupling terms between time and lag. Then, Matched Fourier transform (MFT) and Robust Energy accumulation (REA) are utilized to obtain the chirp rate estimate of the received LFM signals. This estimate is used to compensate the time-quadratic phase term of the GAPCK, and then Fast Fourier transform (FFT) along time-axis is performed to estimate the constant coefficient. Moreover, the asymptotic statistical properties of parameter estimates are derived. The proposed method has low computational complexity and favorite performance under low signal-to-noise ratio (SNR) due to the low-order non-linearity of the GAPCK. Finally, simulated and real data are provided to verify the robustness and effectiveness of the proposed method.

在本文中，我们从有限数量的嘈杂离散时间观测结果中，为多分量线性调频（LFM）信号提出了一种快速且鲁棒的参数估计方法。首先，引入了一种新的内核，称为广义可调参数相关内核（GAPCK），以避免时间和滞后之间的耦合项。然后，利用匹配傅立叶变换（MFT）和鲁棒能量累积（REA）来获得接收到的LFM信号的线性调频率估计。该估计用于补偿GAPCK的时间二次相位项，然后沿时间轴执行快速傅立叶变换（FFT）以估计常数系数。此外，推导了参数估计的渐近统计性质。由于GAPCK的低阶非线性，该方法具有较低的计算复杂度和在低信噪比（SNR）下的最佳性能。最后，提供了仿真数据和真实数据，以验证所提出方法的鲁棒性和有效性。