基于RSPWVD-Hough的多分量LFM信号检测

Multi-component LFM Signal Detection Based on Reassign-Smooth-Pseudo-Wigner-Ville Distribution

由于多分量LFM信号的双线性时频分布(BTFD)存在交叉项的干扰,所以在低信噪比条件下直接在时频平面难于进行检测,采用重排平滑伪魏格纳维尔分布(RSPWVD),能在抑制交叉项的同时提高时频聚集性,并利用Hough变换检测图像中直线的原理,将多分量LFM信号的检测问题转换为在参数空间寻找局部极大值的问题,由于Hough变换运算量较大,提出一种在平滑重排后进行设定门限处理的方法,仿真实验表明,该方法可以在抑制噪声和交叉项的同时大大减少运算量。

Because of the presence of cross-term interference in bilinear time-frequency distribution （BTFD） of multi-component chirp signal, it is difficult to detect the chirp components in a time：frequency plane in low SNR. Using reassign-smoothed -pseudo-wigner-ville（RSPWV） can restrain cross-term and enhance time-frequency concentration, and change the detection problem in multi-component chirp signal to searching local max in parameter space through Hough Transform. The threshold was set after RSPWVD to reduce calculation of Hough Transform. Simulation results show that the method can suppress noise and cross-term interference, and shorten calculation time.