一种基于决策导向图的通信信号调制识别方法

A Modulation Recognition Method of Communication Signal based on Decision Directed Graph

通信信号调制识别在监测系统中具有重要作用,可为后续任务提供参数依据和情报支撑。识别信号的调制样式,是实现信号正确接收解调的前提。针对调制样式识别问题,提出了一种采用决策导向图结合小波变换的方法,更好地完成了信号分类识别。在分类思路方面,引入判决树结构进行分析,并采取决策导向图对支持向量机分类器实现扩展优化,进一步提高了抗噪性能。该方法稳健性好,便于工程实现,并通过仿真试验验证了其识别结果的有效性,识别正确率高且性能较优。

Modulation recognition of communication signal plays an important role in the monitoring system,which can provide parameter basis and information support for subsequent tasks.Identifying the modulation pattern of the signal is the premise to realize the correct reception and demodulation of the signal.Aiming at the problem of modulation pattern recognition,a method using decision oriented graph combined with wavelet transform is proposed to better complete signal classification and recognition.In terms of classification ideas,the decision tree structure is introduced for analysis,and the decision directed graph is used to expand and optimize the support vector machine classifier,which further improves the anti-noise performance.This method has good robustness and is convenient for engineering implementation.Finally,the effectiveness of the recognition results is verified by simulation experiments,and the recognition accuracy is high and the performance is fairly good.