【摘要】 提出了一种基于频域高阶累积量的信号频率和到达角联合估计算法。首先分析阵列信号频域模型 ,然后利用频域数据构造类似于VESPA算法的矩阵束 ,并基于此矩阵束给出频率和到达角联合估计的方法。该算法具有抑制色高斯噪声及实现阵列虚拟扩展的能力 ,并可降低对阵列一致性的要求 ,用M个阵元可以分辨M- 1个空间信号 (ESPRIT类算法 ) ,而计算量可以和基于二阶统计量的算法比拟。更多还原

【Abstract】 In this paper, a new cumulants-based method for the joint estimations of frequency and DOA of spatial signal was discussed. After analysing the model of data in DFT domain, a method to calculate the cumulants matrix pencil of VESPA is proposed, which needs little calculations compared to the direct calculating method. A new algorithm is proposed to estimate the frequency and DOA simultaneously based on the matrix pencil. This method utilized the capability of cumulant of aperture extension and insensitivity to additive colored Gaussian noise. M-1 signals can be processed by this method with M sensors without searching procedure and only a moderate calculations is needed.更多还原