一种有色噪声背景下混合信号的波达方向估计算法

A novel direction of arrival estimation algorithm for uncorrelated and coherent signals with colored noise

提出一种有色噪声背景下,不相关信号和相干信号构成的混合信号的测向算法。在本文算法中,不相关信号和相干信号采用并行处理方式,缩短运算时间。首先,通过对划分的阵列输出矢量子阵构造2组自相关矩阵和互相关矩阵,利用它们构造新的矩阵并进行特征值分解,估计出不相关信号的来波方向;然后利用Toeplitz矩阵的特性,采用空间差分法去除混合信号中的不相关信号和有色噪声;再采用本文提出的改进的矩阵分解法解相干,最后应用ROOT-MUSIC算法对重构矩阵进行波达方向估计。本文所提算法在进行混合信号的波达方向估计问题时充分利用接收信号2组自相关矩阵和互相关矩阵,同时不需要进行谱峰搜索,计算量较小。理论分析和实验仿真结果验证本文算法的有效性和良好的测向性能。

A novel direction of arrival estimation algorithm for uncorrelated and coherent signals with colored noise was proposed. Parallel estimation for uncorrelated and coherent signals was utilized in this algorithm, which helped to reduce the computation time. First, a new matrix was reconstructed by exploiting multiple divisional array output vector submatrices, and the DOAs of uncorrelated signals were obtained through eigen-decomposition of this new matrix. Then, a spatial differencing method was adopted to eliminate the interference of uncorrelated signals and colored noise by taking advantage of Toeplitz characteristic. Moreover, the proposed improved matrix decomposition method was carried out to reconstruct a new matrix. With the ROOT-MUSIC algorithm, the DOAs of coherent signals were achieved. The proposed algorithm makes full use of multiple autocorrelation-matrices and cross correlation-matrices of received signals, which requires no spectral peak searching. Hence, the computational burden is low. Theoretical analysis and simulation results show the effectiveness and performance of the proposed algorithm.