基于离散多项式相位变换和分数阶傅里叶变换的加速目标检测算法

 本文提出了一种离散多项式相位变换(DPT)和分数阶傅里叶变换(FRFT)相结合的LFM检测方法.首先,分析了经典DPT算法中调频分辨率与互相关时延、积累脉冲长度以及积累性能之间的相互关系,指出时延较小的DPT算法与经典DPT算法相比,检测性能提高约3dB,但调频率估计误差大.其次,针对这一问题,提出在DPT基础上,利用FRFT进一步提高调频率估计精度,当输入信噪比大于-11dB时,在分数阶域引入一维牛顿迭代法可提高运算速度;最后,给出了算法复杂度和估计误差的理论分析并用实验结果验证了算法的有效性.

An Accelerating Target Detection Algorithm Based on DPT and Fractional Fourier Transform

The echo of accelerating targets can be approximated as a linear frequency modulated signal which may degrades the MTD′s coherent integration performance.This paper presents an LFM detection method based on both discrete polynomial phase transform(DPT) and the fractional Fourier transform(FRFT).First,the interrelation between frequency resolution,cross-correlation delay,pulse length and integration performance of the classic DPT algorithm are analyzed.The result shows that detection performance of the DPT algorithm with a small time delay can be improved 3dB than classical algorithm.This is helpful to detect accelerating target under low SNR,but causes large parameter estimation error.Second,based on DPT algorithm,the FRFT is applied to further improve the accuracy of chirp rate estimation.When the input signal-to-noise ratio(SNR) is greater than-11 dB,one-dimensional Newton′s iteration method can help to improve the operation speed in the fractional domain.Finally,the algorithm complexity and the theory of estimation error are presented,and the validity of proposed algorithm is verified by simulation results which make the theory perfect.