基于二次灰度形态学滤波的跳频信号盲检测  

Blind Detection Method of FH Signals Based on Twice Gray-Level Morphological Filtering

作　　者：古培 王斌[1] 

机构地区：[1]信息工程大学,河南郑州450001

出　　处：《信息工程大学学报》2015年第1期79-83,128,共6页Journal of Information Engineering University

基　　金：国家自然科学基金资助项目(61201381)

摘　　要：短波复杂电磁环境增加了跳频检测难度,现有图像处理类盲检测法大都需要先预设门限进行二值化,但理想门限较为困难且二值图像忽略了跳频的灰度形态特征,因此针对灰度时频图运用形态学滤波,给出了基于二次灰度形态学滤波的检测算法。首先对时频图频率分量进行灰度形态学滤波,滤除大部分尖锐噪声分量和扫频干扰信号;然后将频率分量去均值,降低定频干扰信号的灰度级,同时保留跳频形态特征;最后对时频图时间分量进行改进顶帽变换,提取跳频信号的二值时频图完成检测。仿真表明,算法能有效克服噪声和干扰信号影响,在大于-10d B时提取较为完整的跳频图案,且算法简单、易于工程实现,为短波跳频信号的盲检测提供了一个新的解决方案。It is difficuh to detect FH signals from the complex electromagnetic environment in HF channels. The existing detection methods by using image processing mostly need binarization images based on the preset threshold, but it＇ s difficult to estimate the threshold, and the gray-scale morphological characteristics of FH signals are ignored in the binary image. Therefore, this paper processes the gray-scale time-frequency spectrogram by using morphological filtering, which gives a new detection method based on twice gray-level morphological filtering. Firstly, the frequency com- ponents of the spectrogram are filtered by morphological filtering, which filters the sharp noise com- ponents and the LMF interference signals. Then it subtracts the mean of frequency components to reduce the gray-scale levels of fixed-frequency interference signals, while retaining the morphological characteristics of FH signals. Finally it extracts the binary spectrogram of FH signals by making the top-hat transformation to the time components of the spectrogram. Simulation shows that the algo- rithm can effectively overcome the noise and interference signals, and extract the complete frequency hopping pattern when SNR is greater than -10dB, which is simpler and easier for project implementation. This paper provides a new solution for the blind detection of FH signals in HF channels.

关 键 词：跳频检测 时频分析 灰度形态学滤波 图像处理 

分 类 号：TN911.72[电子电信—通信与信息系统]