通信信号调制类型的三角矩特征盲识别

针对盲环境下的通信信号调制类型识别困难的问题,提出了一种新的信号盲聚类算法,该算法利用采样信号的瞬时频率作为训练样本,以方向数据统计理论为依据提取分类特征,利用该特征参数在二维特征平面实现了多种不同通信信号的调制类型识别.该方法不仅能实现类间识别,同时也能够实现类内识别.仿真实验结果表明,该算法简单高效,识别率较高,鲁棒性较好,具有较强的实用性和可行性.

Blind identification of a communication signal's modulation type by sampling features of trigonometric moments

This paper presents a new blind clustering algorithm to solve the difficult problem of blind identification of the modulation type of communication signals. The algorithm uses the instantaneous frequencies of sampled signals as training samples and extracts classification features based on the statistical theory of directional data. The classified characteristic is utilized to achieve modulation recognition of various communication signals in a two-dimensional characteristic plane. This method allows identification not only between different types, but also within the same type. Simulation results show that the algorithm is simple, efficient, highly accurate and robust. This makes it both useful and practical.