一种基于协方差阵拟合的解测向模糊算法  

Approach of ambiguity resolution based on covariance matrix fitting

作　　者：刘洪盛[1,2] 肖先赐[2] 

机构地区：[1]电子科技大学通信与信息工程学院,成都610054 [2]电子科技大学电子工程学院,成都610054

出　　处：《计算机应用研究》2009年第9期3293-3295,3299,共4页Application Research of Computers

摘　　要：流形模糊发生的原因是阵列流形上的多个矢量线性相关,子空间类算法无法靠自身解模糊。为解决子空间类算法中的模糊问题,提出了一种新的基于协方差阵拟合的解模糊算法。该方法用预估的导向矢量与入射波功率来拟合协方差阵,用最陡下降法寻找最优的功率参数,认定功率较大的入射波为真实来波。与子空间算法结合,可解决测向模糊问题。数值仿真表明,此方法在阵元数量较多时仍能有效工作。The reason of the manifold ambiguity is the linear dependence of steering vectors on manifold, this kind of ambigui- ty can not be resolved by subspace algorithms themselves. To resolve the manifold ambiguity, this paper proposed an approach based on covariance matrix fitting. The implementation of the approach was to fit the covariance matrix by pre-estimated mani- fold vectors and corresponding powers, then found optimal solution of powers by method of steepest descent. At last, regarded the DOAs with bigger powers as true DOAs. Combining with subspace algorithms, the approach could identify the true DOAs. Simulation verify its effectiveness and robustness when there are more sensors in array .

关 键 词：阵列 测向 模糊 协方差阵 拟合 子空间 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]