摘要
有效的方位特征获取对构建智能来波方位估计模型具有十分重要的意义。该文在分析阵列接收信号相关函数的基础上,首次提出利用相邻阵元信号互相关函数的相角作为来波方位特征。与常用的协方差矩阵上三角特征相比,剔除了与来波方位无关的幅度信息和冗余的方位特征信息,在不损失有效方位信息的基础上使特征维数得到极大地降低。实验结果表明,利用相角特征构建的RBF神经网络的结构更简洁,泛化性能更好,来波方位估计精度高,实时性好,具有广阔的工程应用价值。
Effective feature extraction is very important when building the smart DOA estimation model. Based on analyzing the correlation function of the array signal, this paper firstly presents using the angles of contiguous array signal＇s correlation function for DOA estimation purpose instead of common used upper triangular half of the covariance matrix, which eliminates the irrelevant magnitude information and redundant direction characteristic. Therefore the feature dimension is largely reduced without losing any DOA information. Experimental results show that the performance of RBF neural network using proposed Phase-feature is superior to the common used upper triangular half of the covariance matrix in terms of neural network size, generalization, estimation precision and real-time performance, so it has a broad application value.
出处
《电子与信息学报》
EI
CSCD
北大核心
2009年第12期2926-2930,共5页
Journal of Electronics & Information Technology
基金
国家自然科学基金(60672161)资助课题
