论文部分内容阅读
本文导出了在复信号输入条件下自适应线性组合器的主要性能公式以及输入相关矩阵正交变换的特性。在此基础上分析了根据梯度搜索性能表面的两种主要方法——牛顿法和最陡下降法以及梯度估值误差对自适应处理性能的影响。最后,作寿应用例子给出了复LMS自适应算法的有关公式。
In this paper, the main performance formulas of adaptive linear combiner and the characteristics of orthogonal transform of input correlation matrix are derived under the condition of complex signal input. Based on this, two main methods of searching for the performance surface based on gradient - Newton method and steepest descent method and the influence of gradient estimation error on adaptive processing performance are analyzed. Finally, the life application example gives the relevant formula of complex LMS adaptive algorithm.