四元数信号模型保持了阵元各分量之间固有的正交特性,因而针对极化敏感阵列信号处理基于四元数的方法具有比常规基于复数的长矢量方法更优的性能。将最大信号干扰噪声比准则应用于波束形成求得四元数最优权向量,进而完成了极化敏感阵列的滤波。与长矢量方法相比在具有相同的滤波性能情况下,四元数方法减少了一半的数据存储单元,降低了除法计算复杂度,从而提高了算法的处理速度。同时四元数的四维超复数正交结构提高了阵列对指向误差的稳健特性,计算机仿真结果验证了文中方法的有效性。
Quaternion signal model can keep the intrinsical orthogonality of the polarization sensitive array(PSA), so the processing method based on quaternion outperforms the performance of traditional long vector method based on complexor for PSA. The quaternion optimal weight vector is obtained through the maximum signal interference noise ratio criterion and it is used to complete the filtering of PSA in this paper. The quaternion method can achieve the same filtering performance as the long vector method. Also, the computational burden is reduced and only a half of the memory space is required. At the same time, four dimensional orthogonal structures of quaternion can enhance the robustness when there exists array pointing errors. Simulation results are given to demonstrate the effectiveness of the proposed method.