基于容积卡尔曼滤波的约束恒模波束形成算法  

Constrained constant modulus beamforming based on cubature Kalman filter

作　　者：沈锋[1] 宋金阳[1] 

机构地区：[1]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001

出　　处：《系统工程与电子技术》2016年第12期2700-2704,共5页Systems Engineering and Electronics

基　　金：国家自然科学基金(61374208);中央高校基本科研业务费专项基金(HEUCFX41310)资助课题

摘　　要：提出一种基于容积卡尔曼滤波的线性约束恒模波束形成算法。首先使用伪观测法将恒模代价函数和约束条件写成状态观测方程,之后利用容积卡尔曼滤波算法来求解以上非线性滤波问题。所提方法能够避免常规算法对模型的近似处理和特征值分散效应对波束形成器输出性能的影响,因此对干扰和噪声有更强的抑制能力。仿真结果表明,本算法相比随机梯度法和递归最小二乘法具有更快的收敛速度和更高的输出信干噪比,在非平稳环境下,能够迅速调整权值收敛到最优解。A constrained constant modulus adaptive beamforming method based on cubature Kalman filter is proposed. Firstly, the state space model of the constrained constant modulus criterion is formulated with the pseudo-observation method. Then the cubature Kalman filter is utilized to solve the non-linear filtering prob- lem. This algorithm avoids the approximation procedure and the high eigenvalue spread effect which exists in conventional methods, and hence it exhibits better interference and noise suppression capacity. Simulation re- sults demonstrate that the proposed method converges faster than both the stochastic gradient method and the re- cursive least square method and possesses higher output signal-to-interference-plus-noise ratio. Additionally, it can ad- just its weight to the optimal result quickly under the non-stationary environment.

关 键 词：自适应波束形成 恒模准则 线性约束 伪观测法 容积卡尔曼滤波 

分 类 号：TN911.7[电子电信—通信与信息系统]