The traditional time⁃frequency analysis has poor anti⁃noise performance,severe cross interference and higher complexity. To overcome these disadvantages, this paper proposes a new algorithm based on both time⁃frequency analysis and morphology pre⁃processing,which has the function of signal detection,parameter estimation and multi⁃access mode recognition.Firstly,a filtering method based on mathematical morphology is presented to eliminate the effects caused by noise floor fluctuation in satellite communication. Then,a new time⁃frequency spectrogram generation method is designed,and the time⁃frequency spectrogram is treated as the two⁃dimen⁃sional image to finish morphology pre⁃processing.Last but not least,the expected function can be realized by analyzing the processed time⁃frequency spectrogram.The algorithm is efficient and easy to implement with low computation complexity.The theoretical analysis and simulation results prove its availability.%为克服传统时频分析对噪声敏感、交叉干扰严重及运算量大等缺点，提出一种新的基于时频分析和形态学预处理的算法，该算法能够实现信号检测、参数估计及多址方式识别的功能。采用数学形态滤波方法消除卫星频带噪底起伏带来的影响；设计了一种新的时频谱图生成方法，将生成的时频图作为二维图像进行形态学预处理；对处理后的谱图进行分析，检测信号是否存在，对信号参数作出初步估计并对信号的多址方式进行识别。该算法简单易实现，噪声对检测的影响较小，理论分析和仿真结果证明了算法的有效性。
展开▼
