多分量相移键控信号调制识别技术

Modulation Classification of Multi-Component PSK Signals

为解决单通道接收机在多信号环境下的调制识别问题,提出了一种不经信号分离直接提取各信号分量的特征不变量以实现多个相移键控信号的调制识别算法。首先给出了多信号特征不变量提取原则,进而提出了融合二、四阶循环累积量的特征不变量提取方法,推导了所提特征不变量的理论值,最后设计了适合于这里模型的分类器。仿真实验表明:该算法能不受干扰地提取各分量的特征不变量,进而分类器能有效地识别调制组合内任意的双分量相移键控信号。同等实验条件下,所提算法的识别性能优于现有的基于四阶循环累积量的算法。

To solve the modulation classification of single-channel receiver in a crowded electromagnetic spectrum,a new modulation classification algorithm is proposed,which could without blind signal separation,directly extract invariant features of each signal component,thus to realize modulation classification.Firstly,the principle for extracting invariant features of multi-signal is given;then the method in combination of invariant features of the 2nd and 4th cyclic cumulants proposed and theoretical value of the extracted invariant feature deduced,finally the suitable classifier for the model in this paper is designed.Simulation results indicate that the proposed algorithm could extract invariant feature of each independent signal without disturbance,and also the classifier could effectively recognize the arbitrarily-combined two-component MPSK signal,and under the same experiment condition,the proposed algorithm has better performance than the existing algorithm based on 4th cyclic cumulant.