无线电监测中调制信号特征提取算法

Modulation Signal Feature Extraction Algorithm in Radio Monitoring

针对传统调制信号特征提取算法在噪声环境下存在识别准确度低、分类效果差等问题,基于已有的调制信号处理方法,提出一种新的无线电监测中调制信号特征提取算法。首先构建无线电监测中各类调制信号的数学模型,以此为基础通过仿真得到信号瞬时幅值、瞬时相位及瞬时频率的特征。分析当前信号调制方式识别各类算法的优缺点,采用小波变换完成调制信号的降噪处理与突变边界特征提取算法的设计,利用零中心归一化瞬时幅度的谱密度最大特征提取算法以及核判别分析算法对各类调制信号进行逐层提取,实现了各类调制信号的完整分类与提取,提升了噪声环境影响下的特征信号提取精度、且分类效果较好,为无线电监测中调制信号特征提取提供了有利科学依据。

Aiming at the problem that the traditional modulation signal feature extraction algorithm has low recognition accuracy and poor classification effect in noisy environment,based on the existing modulation signal processing method,a new feature extraction algorithm for modulation signals in radio monitoring is proposed.The mathematical models of all kinds of modulation signals in radio monitoring are constructed firstly,based on which the characteristics of instantaneous amplitude,instantaneous phase and instantaneous frequency are obtained by simulation.The advantages and disadvantages of various algorithms are analyzed for the recognition of the current signal modulation mode,using the wavelet transform to complete the noise reduction processing of the modulation signal and the design of the algorithm for the feature extraction of the abrupt boundary,use the algorithm for the maximum feature extraction of the spectral density of the zero center normalized instantaneous amplitude and the kernel discriminant analysis algorithm to extract all kinds of modulation signals layer by layer.The complete classification and improvement of all kinds of modulation signals are realized.The extraction accuracy of the feature signal under the influence of noise and the classification effect are improved,which provides a favorable scientific basis for the feature extraction of modulation signals in radio monitoring.