论文部分内容阅读
针对传统的时域、频域特征参数在干扰识别中识别率不高,识别效果不稳定等问题,文章提出了基于循环谱奇异向量的干扰识别方法。在研究干扰信号循环谱的基础上对循环谱采用矩阵奇异分解方法提取部分左右奇异向量,最后采用概率神经网络分类器实现干扰信号的分类识别。仿真结果表明,该方法整体效果较好,尤其在较低干信比条件下明显优于传统识别方法,验证了该方法的有效性。
Aiming at the problems of traditional time domain and frequency domain feature parameters, such as low recognition rate and unstable recognition effect, the paper proposes an interference identification method based on singular vector of cyclic spectrum. On the basis of studying the cyclic spectrum of the interfering signal, a matrix singular decomposition method is applied to the cyclic spectrum to extract some left and right singular vectors. Finally, the classification and identification of the interfering signal are carried out by using the probabilistic neural network classifier. The simulation results show that the proposed method is better than the traditional method, especially under the condition of lower dry-letter ratio, which verifies the effectiveness of the proposed method.