面向矢量声信号的实值EB-ESPRIT算法  

Real-valued EB-ESPRIT Algorithm for Acoustic Vector Signals

作　　者：石超宇 王志强[1] SHI Chaoyu;WANG Zhiqiang(School of Communication and Information Engineering,Shanghai University,Shanghai 200444,China)

机构地区：[1]上海大学通信与信息工程学院,上海200444

出　　处：《电声技术》2020年第6期87-90,共4页Audio Engineering

摘　　要：到达方向(DOA)是声学场景分析中非常重要的参数,而准确、高效的DOA估计有助于后续信号处理,尤其是实时性较强的应用场景。因此,提出面向矢量声信号的实值EB-ESPRIT算法,将球麦克风阵列的DOA估计方法应用于矢量声信号,用于降低计算复杂度。根据矢量声信号导向矢量与低阶球谐函数的一致性关系,将矢量声信号转换为球谐信号,使用球谐特征波束对声源DOA进行求解。实验表明,提出的方法可以在保证较好估计精度的同时有效降低计算开销。The Direction of Arrival(DOA)is a very important parameter in the analysis of acoustic scenes.Accurate and efficient DOA estimation is helpful for subsequent signal processing,especially in real-time application scenarios.In this paper,the real-valued EBESPRIT algorithm for vector acoustic signals is proposed,and the DOA estimation method of the spherical microphone array is applied to vector acoustic signals to reduce the computational complexity.Specifically,the vector acoustic signal is converted into a spherical harmonic signal according to the consistent relationship between the vector acoustic signal steering vector and the low-order spherical harmonic function,and the spherical harmonic characteristic beam is used to solve the sound source DOA.Experiments show that the method proposed in this paper can effectively reduce the computational cost while ensuring better estimation accuracy.

关 键 词：DOA ESPRIT 特征空间算法 声矢量传感器 

分 类 号：TN912.16[电子电信—通信与信息系统]