基于两步最小均方的宽带波束形成算法

An Broadband Beamforming Algorithm Based on Two-Step LMS

针对空时二维滤波器结构,提出了修正两步最小均方波束形成算法,通过修正每次迭代运算中的期望阵列响应幅度和方向矩阵,显著地减少了矩阵运算量,达到快速准确求解期望方向图的相位和相应权矢量的目的。该算法在不改变天线阵元数的情况下,比传统空域滤波增加了时域处理维数,提高了波束形成质量,并利用了期望方向图相位的随机性,通过迭代更新期望方向图的相位和相应权矢量,实现了方向图的旁瓣控制,理论上可以形成任意控制的旁瓣波束,满足不同应用场合的需求。经计算机仿真实验,算法可以很好的收敛到期望的方向图,验证了该算法的有效性。该算法可广泛应用于多天线的GNSS导航接收机或抗干扰接收机中,在提高接收机在复杂环境下的可用性和灵敏度上具有重要的意义。

In this paper, a modified two-step LMS space-time filtering based beamforming algorithm is proposed. By modifying the desired amplitude response and steering matrix in each iteration, the complexity of the matrix computation is significantly reduced, achieving faster and more accurate calculation for both the phase of the desired pattern and the weight vector. Using the proposed method in both temporal and spatial domain, the quality of beam forming can be improved without additional antennas. By making use of the randomicity of the desired pattern phase, this method can be used to iterate and update the phase of desired pattern and weigh vector to control the side-lobe, so it has the potential to implement any desired pattern and to be used in a wide range of applications. Computer simulations verified the validity of the method, which has good convergence to the desired pattern. This method can be widely used in GNSS navigation receivers or the anti-jamming receivers with antenna array, considerably improving the usability and sensitivity in complex environments.