基于流形学习的LFM信号模糊函数特征提取与识别研究

现代多功能雷达向多功能、多用途方向发展,具有多种工作体制与工作状态,同时,雷达采用复杂的波形设计。现代电磁环境复杂,杂波、噪声干扰严重。上述原因使得雷达信号的特征提取与识别变得越来越困难。传统雷达信号的特征提取与识别方法分别是针对雷达信号特征提取过程中存在的某一个或几个方面的问题提出的,各有优点和不足,并不适用于所有情况下的所有信号。提出用流行学习对线性调频(LFM)信号模糊函数进行特征提取,并对其进行了仿真与分析。

Research into The Extraction and Recognition of LFM Signal Ambiguity Function Feature Based on Manifold Learning

Modern multi-function radar is developed more and more multi-function and multi-pur- pose, has multiple operation systems and states, and adopts complicated waveform design. Mordern electromagnetism environment is complicated, and the clutter/noise jamming is serious. The above reasons make the feature extraction and recognition more and more difficult. Traditional feature ex- traction and recognition methods of radar signal are usually put forward aiming at a certain or sev- eral problems existed in feature extraction process of radar signal, have advantages and disadvanta- ges seperately,are not adapted to all signals in all states. This paper presents the manifold learning to perform feature extraction to the ambiguity function of linear frequency modulation signal, which has catholicity,and performs simulation and analysis to the method.