基于切片熵权的WHT多LFM信号识别方法

为提高在低信噪比与先验信息不足条件下对线性调频(LFM)信号识别能力,借鉴信息论中的熵权法改进WHT(Wigner-Hough Transform),提出了一种基于切片熵权的WHTE(Wigner-Hough Transform based on Entropy)算法。推导出LFM信号的WHT与对应特性,将WHT变换域内极半径和角度切片的熵值来转换为权重因子,进而对每个切片进行加权处理,采用双层权重以弱化噪声与干扰项的影响,并推导出LFM信号与高斯白噪声在WHT维度内不同假设条件下的概率密度分布函数,构建了对于LFM信号WHT后恒虚警检测的完备流程。通过理论分析与公式推导论证了算法的可行性,并与WHT、分数阶傅里叶变换与周期WHT算法的仿真对比,验证了算法的有效性,凸显WHTE算法能够在强噪声背景下与没有先验支撑时实现对LFM信号的良好检测。

Recognition method of multiple LFM signals with WHTbased on entropy of slice

In order to improve capability of recognizing linear frequency modulated (LFM) signal under low signal-to-noise ratio and insufficient prior information,Wigner-Hough transform (WHT) is improved with entropy method of information theory,and an algorithm of Wigner-Hough transform based on entropy (WHTE) of slice is proposed.The WHT and corresponding features of LFM signal are derived;the entropy of polar radius and angle slices in WHT's transform domain is transformed to weight factor,and then each slice can be weighted.Double-deck weight is used to weaken the influence of noise and jamming term,and probability density distribution function of LFM signal and Gaussian white noise under different assumptions in WHT dimensionality is deduced;constant false alarm rate inspection's complete flow of LFM signal after WHT is established.The feasibility of algorithm is verified via theoretical analysis and formula derivation,and the effectiveness of algorithm is proved according to comparisons with WHT,fractional Fourier transform and periodic WHT,which shows fine detection of LFM signal by WHTE algorithm under strong noise and insufficient prior information.