脉冲噪声环境中基于clipping方法的鲁棒LPFT及其重排算法

摘要

In this paper, a new clipping technique based robust LPFT is proposed to process time-varying signals in the impulsive noise. First, the clipping technique is used to suppress the impulsive noise to achieve good time-frequency representation for the time-varying signal. Then, the reassignment method is combined with the LPFT to improve the time-frequency concentration in the time-frequency domain. Simulations show that, compared with the median filter based robust LPFT, the clipping technique based robust LPFT can also provide good representation for signals in the impulsive noise, with smaller MSE and less computational complexity. Moreover, the reassignment method is combined with the clipping technique based robust LPFT to effectively improve the time-frequency concentration. Therefore, the clipping technique based robust LPFT and its reassignment method is an effective method to process signals corrupted by the impulsive noise, as well as to achieve improved time-frequency concentration.%为了更好地处理脉冲噪声环境中的时变信号,本文提出了基于clipping方法的鲁棒局部多项式傅里叶变换(LPFT)及其重排算法.首先利用clipping方法对信号中掺杂的脉冲噪声进行抑制,得到较好的信号时频分布表示,然后将重排算法与该鲁棒LPFT相结合,以提高信号的时频聚集性.通过实验仿真可以看出,与基于中值滤波器的鲁棒LPFT相比,基于clipping方法的鲁棒LPFT同样能对被脉冲噪声干扰的信号给出较好的时频表示,而且其瞬时频率估计的最小均方误差(MSE)较低,计算量较小.并且,本文在基于clipping方法的鲁棒LPFT对掺杂脉冲噪声的信号进行处理的基础上,利用重排算法与其结合,有效增强了信号的时频聚集性.因此基于clipping方法的鲁棒LPFT及其重排算法是一种高效的处理脉冲噪声干扰信号及提高信号时频聚集性的方法.