一种有效的通信电台综合识别技术

Study on Techniques of Processing Features and Recognition of Radio Transmitters

由于通信电台的"指纹"特征很难获取,电台识别面临巨大的挑战,本文采用综合识别的方法,对提取信号的多个特征进行特征相关性和特征重要性处理,剔除有强相关性的特征,并对特征属性在识别过程中利用获取的特征重要性进行加权。通过对实际空间传播的FM电台信号的综合识别实验表明在现有的信号处理技术条件下,通过对提取的特征进行有效分析与处理,能够提高电台识别的正确率,验证了对提取特征进行相应处理的有效性和必要性。

After analyzing the correlation among features extracted from radio transmitter signals and processing their attributes signifi-cance, some strong correlation features are removed and attributes weighted features are obtained when using classification algorithm. The experimental results of radio transmitter identification through FM signals show that accuracy of recognition is increased after analysis on the efficient features, and prove the effectiveness and necessary of correspond processing the extractin...