软件无线电中调制识别算法的研究

在非合作通信体制下，通信信号的调制样式识别已经越来越占据重要作用。以往文献中可以参考到的算法一般要求信噪比比较高，一般要求在10db以上。该文基于软件无线电体系结构，在低信噪比情况下对数字通信信号的识别算法作了一些改进。通过在参数提取时适当选取预处理滤波器和合理对相位进行修正来降低噪声的影响，并选择适当的判别参数来提高算法的鲁棒性，使得算法在信噪比为3db的环境下对常用数字调制样式识别的仿真结果达到90％以上。

Classification of Digital Signal Modulations in Software Defined Radio (SDR)

The importance of the classification of signal modulation is more and more apparent in the uncooperative communication systems.Many algorithms th at have been proposed require high SNR approximately 10 db.Therefore,this paper discusses some improved methods in the extraction of the parameters such as sel ecting preprocessing filter,phase compensating and choosing suitable identificat ion features to make the algorithm more robust under the low SNR. And the simula tion shows that the identification rate can reach 90% in a SNR of 3Db.