基于稀疏近似最小方差的宽带DOA估计算法

Wideband Signal DOA Estimation Based on Sparse Asymptotic Minimum Variance

提出了一种基于稀疏近似最小方差的宽带波达方向（DOA）估计算法，有效解决当前宽带波达方向估计方法分辨率较低、相干干扰条件下DOA估计误差大等问题。该算法在空域稀疏模型的基础上利用宽带相干信号子空间算法，将宽带信号聚焦到固定频点处，再以近似最小方差准则进行迭代，实现高分辨DOA估计。该算法得到的空间谱具有稀疏度高和副瓣低的特点，且与CSM算法相比无需任何先验信息。通过仿真计算表明，该算法的方位分辨能力较宽带传统DOA估计算法提升30％，并具有分辨相干信号能力。通过某海试实测数据处理得到的时间历程图有效验证该算法性能。

A wideband signals direction-of-arrival （ DOA ） algorithm based on sparse asymptotic minimum variance（ SAMV） is presen- ted in this paper,which can effectively improve spatial resolution and anti-coherent-interference performance. Based on spatial sparse signal model, the algorithm transforms wideband signal to narrowband, using coherent signal subspace method（CSM）, and then iter- ates with asymptotic minimum variance （AMV） approach to carry out high-resoluition DOA estimation. The spatial spectrums imple- mented by this algorithm have the specialty of high sparse and extremely low side lobe. What＇s more, the prior information isn＇t re- quired in this algorithm. Computer simulations show that, this method improves 30% spatial resolution compared to traditional DOA algorithm, and coherent interference resistance. Bearing-time recording （BTR） results of a sea trial also demonstrate the efficiency of the algorithm.