传统的稀疏恢复DOA估计方法需要建立多维网格进行联合优化，计算量大且存在网格失配问题。本项目研究基于球上的超分辨理论，研究不同快拍、信号环境下的球谐波域DOA估计方法，旨在实现共形阵列的高精度快速DOA估计。主要研究内容包括：(1)利用对偶多项式理论，研究多测量快拍下的球上超分辨方法，突破现有方法单快拍测量的限制，实现基于超分辨的球谐波域窄带DOA估计。(2) 通过构建对偶多项式对各频点阵列数据的映射，研究多信号模型下的球上超分辨方法，克服现有方法限于单一模型的不足，实现基于超分辨的球谐波域宽带DOA估计。 (3)构建球谐波域co-array虚拟阵列，利用不同快拍和模型下的球上超分辨方法，研究欠定情况下共形阵列窄带与宽带DOA估计方法。上述方法能够避免求解多维网格联合优化问题和网格失配，增加阵列自由度，提升DOA估计性能。本项目研究成果对于设计高性能共形阵列测向系统具有理论意义和实用价值。

The traditional sparse recovery DOA estimation methods need to construct multidimensional grid and solve joint optimization program which cause high computing burden and grid mismatch problem. This project is concerned the spherical harmonic domain DOA estimation based on the super resolution on the sphere. The main research content of the project are as followed: (1) the narrowband DOA estimation with multiple snapshots in spherical harmonic domain based on the super resolution is studied.(2) the wideband DOA estimation in spherical harmonic domain based on the super resolution is proposed. 3) By utilizing the super resolution method for different snapshots and models and constructing the co-array virtual array in spherical harmonic domain, we study the narrowband and wideband underdetermined DOA estimation for the conformal array. These methods can avoid the high computational burden and grid mismatch problem, enhance the freedom degrees and improve the performance of DOA estimation. The results of this project will be helpful for designing high performance DOA estimation with conformal arrays.