稀疏圆阵的解相干求根MUSIC算法

Sparse UCA Decorrelation Root-MUSIC Algorithm

针对稀疏圆阵的波达方向估计问题,提出了解相干求根MUSIC算法(Sparse UCA Decorrelation Root-MUSIC,SDR)。通过改进传统的波束变换方法,进行相位校正,并在波束域进行误差补偿,得到具有共轭对称结构的波束域导向矢量。在波束域进行前后向平均处理和使用求根MUSIC算法,实现多组相干源的解相干,且避免了谱搜索,减少了运算量。平均处理增加了数据量,算法在低信噪比和低快拍数情况下有更好的估计性能。计算机仿真表明,本算法适用于稀疏圆阵对相干源的DOA估计而且有较好的估计性能。

The sparse UCA decorrelation root-MUSIC algorithm（SDR） is proposed to estimate the DOA of coherent signals of sparse uniform circular array. It improves traditional beamspace transform method by calibrating the phase of steering vector and eliminating the phase error. The steering vector of beamspace do- main is centric hermitian, so forward-backward average method and root-MUSIC can be applied. In this way, the computational burden is reduced by avoiding the search for the highest peaks and the decorrelation is achieved. Simulation results show that this method works well in sparse UCA even under low-SNR or low- snap situation.