基于稳定分布模型的自适应阵列信号处理新方法

New methods of adaptive array signal processing based onstable distribution model

传统方法常对阵列信号处理所研究的噪声采用高斯分布的模型进行描述,但当噪声存在显著的尖峰时,往往不能得到满意的结果。利用稳定分布建模实际中所遇到的具有较大脉冲特性的随机噪声,综述了稳定分布模型下的信源定位、波达方向估计、波束形成等阵列信号处理方法,并利用分数低阶统计量提出了几种较有韧性的阵列信号处理新方法。仿真表明它们在高斯和低阶α稳定分布噪声条件下,具有良好韧性与有效性。

Conventional methods of array signal processing are based on Gaussian assumption, which can not lead to a good performance. Alpha stable distribution can better model the impulsive random processes in physical applications. The existing adaptive array signal processing methods under alpha stable noise conditions are summarized, and several new robust methods are proposed. The effectiveness of the new algorithms is proved by computer simulations.