前后向平均匹配场处理

A Matched Field Processor Using Forward /Backward Averaging Algorithm

为了提高匹配场被动定位的稳健性，提出了一种基于白噪声增益约束的前后向平均匹配场处理器。算法不仅使用了白噪声增益约束了权向量，而且针对垂直线列阵在空间上均匀分布的特点，在估计协方差矩阵时使用了前后向平均算法，因此提高了目标检测和定位的性能。同时由于前后向平均算法在估计协方差矩阵时能够把复数运算转化为实数运算，大量减小了运算量。为了验证算法的性能，使用仿真数据和海试数据进行了处理分析。结果表明，和前向平均匹配场处理器相比，该算法的输出信干比提高了约1？5～5 dB，峰值背景比提高了约1～4 dB。

In order to enhance the robustness of the passive location of the matched field processor, we propose a matched field processor that uses the forward/backward （ FB） averaging algorithm and the white noise gains con-straint. The processor not only restrains the weight vector of the matched field with white noise gains but also esti-mates the sample covariance matrix with the FB averaging algorithm that utilizes the spatially uniform distribution characteristics of the vertical linear array, thus enhancing the target detection and location of the matched field. The FB averaging algorithm can also transform complex number operation into real number operation when estimating the covariance matrix, thus greatly reducing the computational complexity and the computational volume. Finally we use the simulation data and sea test data to locate the source. The location results, given in Figs. 2 through 6, and their comparison show preliminarily that the output signal to interference ratio increases by about 1？5 to 5 dB, and that the peak to background ratio increases by about 1 to 4 dB.