基于小波和分形理论的调制信号特征提取方法研究

本文针对通信信号这种非平稳的、信噪比(SNR)变化范围较大的信号,应用小波分析和分形理论,提出了有效的特征提取方法,所提取的分形特征包含了区别不同调制类型所需的幅度、频率和相位等主要信息。同时这种特征具有较好的抗干扰特性,基于这种特征的分类器的设计简单、高效。计算机模拟结果表明此方法的有效性。

A Method of feature extraction for Modulated Signals Based Wavelet and Fractal Theory

Based on wavelet analysis and fractal theory, a new efficient method of feature extraction is presented with the purpose of nonstationary and varied communication signals in wide range of signal noise ratio in this paper, the fractal features extracted contain the changing rules of magnitude, frequency and phase of signals which distinguish various modulation modes. Moreover, these features are insensitive to noises interfering and the classifier design is very simple based on them. The result of computer simulation has proved its good performance.