一种基于对角加载的降维多级维纳滤波实现方法

A Kind of Multistage Wiener Filter Realization Based on Diagonal Loading

  • 摘要: 文章以GPS信号抗干扰为背景,应用多级维纳滤波的抗干扰方式。首先分析了经过多级维纳滤波前向迭代后的降维数据矢量的协方差矩阵,给出以特征值和特征矢量表示的最优权值。最优权值的表示式表明在滤波器维数超过最优时高阶的特征值对最优权值影响较大。随后通过仿真试验证明了在滤波器维数超过最优时输出信干噪比会下降。通过引入对角加载量并适当选择加载量功率的大小可以抵消滤波器维数过估计带来的影响,试验表明在对角加载量与白噪声方差相等时可以获得较好的输出信干噪比。文章从空频响应的角度试验了对角加载对于多级维纳滤波算法的意义。最后仿真试验了基于对角加载多级维纳滤波实现方法的抗干扰性能。

     

    Abstract: The paper is aiming at interference elimination of GPS signal, using the method of multistage wiener filter. Firstly, the covariance of the reduced rank data which passed the front part of multistage wiener filter is analyzed, which educe the optimized weight denoted by eigenvalue and eigenvector. The donation of optimized weight indicate that when the filter rank exceed it’s best, the higher rank eigenvalue and eigenvector infect a lot to the weight. Then it is proved by simulation that when the filter rank exceed it’s best, the SINR of output will drop. The effect of the filter rank over estimate can be diminished by introducing a diagonal constant and properly setting the constant,it is certified that a preferably output SINR can be obtained when setting the diagonal constant equals to white noise variance. The significance of diagonal loading to multistage wiener filter is certified through simulating the spacial-frequency response. At last, the anti-jam capability of multistage wiener filter realization based on diagonal loading simulated and certified.

     

/

返回文章
返回