中文摘要：

针对多分量线性调频信号WVD（Wigner-Ville Distribution）检测中的交叉项问题,提出一种交叉项去除方法.利用自项和交叉项的频率特性,将WVD时频矩阵进行坐标旋转,在变换域上滤波去除交叉项.理论推导了矩阵旋转变换公式以及旋转后自项和交叉项的表达式,并且针对缓慢震荡的交叉项受滤波器性能限制不能完全去除的情况,提出能量加权的方法进行改进.仿真和实验结果验证,该方法不仅能够去除交叉项,且不会降低分辨率.

英文摘要：

A method of cross-term removal for Wigner-Ville distribution is proposed to the detection of multi-component LFM signals.For the frequency characteristics of auto terms and cross terms,it removes cross terms by filtering WVD time-frequency matrix in the transform domain.The matrix transformation formula and the expression of all terms after transformation are given.And in view of the performance of filter,an energy weighted method is put forward to improve the slow oscillating cross term.Simulation and experimental results demonstrate that,it can eliminate the cross term with no reduction of time-frequency resolution.