摘要
基于信号空间时频分布的到达角(DOA)估计算法利用Cohen类分布构造空间时频分布矩阵D_(xx)(t,f)代替传统的阵列协方差矩阵R_(xx),信源数目判断是阵列测向中的重要课题,在基于空间时频分布的DOA算法基础上,对阵列观测结果的空间时频分布矩阵进行模式识别,通过搜索来波在时频平面的信号项进行聚类,可以在信号源数目多于阵元数时判断来波数量。
Based on the spatial time-frequency distribution of signals direction of arrival method uses the Cohen class time-frequency distribution,the spatial time-frequency distribution matrix D_(XX)(t,f) is constructed to replace the traditional correlation matrix R_(XX).Source number decision is an important problem in sensor array processing.Pattern recognition is used to spatial time-frequency distribution matrix of observed data of sensor array.Cluster is performed by searching auto-terms of time-frequency plain.The method can get source number at the situation of source numbers more than array sensors.
出处
《通信技术》
2006年第S1期25-27,共3页
Communications Technology
关键词
空间时频分析
波达方向
模式识别
非平稳信号
spatial time-frequency distribution
DOA
pattern recognition
nonstationary signals
