论文部分内容阅读
特征分析是雷达信号分选识别的基础,利用稀疏分解思想对新体制雷达信号进行特征提取是一个新的研究方向。本文以分数阶Fourier变换的核函数作为稀疏分解的chirp基函数,将具有相近特征参数的chirp基函数构成基函数族用于稀疏分量提取,推导了在分数阶Fourier域基于匹配跟踪的chirp基函数族稀疏分解公式,然后利用chirp基稀疏分量的调频率和初始频率构成特征参数序列,将雷达信号脉冲分成5大类进行分选和识别,仿真分析验证了推导结果的有效性。结果表明对于具有线性或曲线时频特征的雷达信号在信噪比为-3dB,采样频率为500 MHz,观测时间为2μs,调频率不超过100MHz/μs时,仍然具有95%的正确分选概率。
Feature analysis is the basis of radar signal classification and recognition. It is a new research direction to use the sparse decomposition idea to extract the features of the radar signal of the new system. In this paper, the kernel function of fractional Fourier transform is used as the chirp basis function of sparse decomposition, and the chirp basis functions with similar characteristic parameters are used as the basis functions for the sparse component extraction. The chirp basis function based on matching tracking in fractional Fourier domain Then by using the chirp-based sparse components’ modulation frequency and initial frequency, the radar signal pulses are divided into five categories for classification and identification. The simulation results verify the validity of the derivation results. The results show that for the radar signal with linear or curved time-frequency characteristics, the signal-to-noise ratio is -3dB, the sampling frequency is 500 MHz, the observation time is 2μs and the modulation frequency does not exceed 100MHz / μs, .