一种基于判决树的无线信号调制模式自动识别  

A Decision Tree Based Automatic Recognition on Modulation Modes of Wireless Signals

作　　者：彭召敏 黄明[1] 徐泽琨 丁照雨 汪弈舟 PENG Zhao-Min;HUANG Ming;XU Ze-kun;DING Zhao-yu;WANG Yi-zhou(North China University of Technology,Beijing 100144,China)

机构地区：[1]北方工业大学,北京100144

出　　处：《工业技术创新》2018年第2期86-89,共4页Industrial Technology Innovation

基　　金：北方工业大学电子信息工程学院;北京市大学生科学研究与创业行动计划项目的资助与支持

摘　　要：基于判决树系统对信号特征量提取实现了无线信号自动识别。以2PSK、4PSK、2FSK和4FSK四种调制方式为例,引入三大判决变量,其中瞬时幅度谱密度最大值用于PSK和FSK信号的初步识别;瞬时相位非线性分量绝对值的标准差用于2PSK和4PSK信号的细分;瞬时频率绝对值的标准差用于2FSK和4FSK信号的细分。仿真试验表明,判决树算法可以实现6~25信噪比范围内调制模式的自动、快速、高效识别,具有一定的可扩展性,在应急救援通信等领域中具有较高的实用价值。A decision tree algorithm is proposed for automatic recognition on modulation modes of wireless signals.4 modulation modes,namely,2PSK,4PSK,2FSK and 4FSK are taken as an example to introduce 3 decision variables,where,the maximum of transient amplitude spectrum density is used for the preliminary identification between PSK and FSK;standard deviation of absolute value of nonlinear component of transient phase is used for the subdivision between 2PSK and 4PSK;standard deviation of absolute value of transient frequency is used for the subdivision between 2FSK and 4FSK.Simulations show that the proposed decision tree algorithm can realize automatic,fast and efficient identification of modulation modes in the range of 6~25 signal-to-noise ratio,and has certain extensibility.It also has high practical value in the fields including emergency rescue communication.

关 键 词：判决树 模糊识别 信号特征量 信号处理 调制模式 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]