摘要
为有效克服期望信号方向存在指向误差而导致的阵列流形向量失配的问题,提出了一种基于支持向量回归机的波束形成方法.该方法在分析线性约束最小方差波束形成器的基础上,将支持向量机的损失函数引入到线性约束最小方差波束形成器的最优化问题中,从而使基于结构风险最小化原理的支持向量回归机算法与波束形成算法相结合.通过MATLAB仿真实验,在没有失配的理想情况和期望信号存在方向向量失配的情况下,选取不同的支持向量机参数以及信噪比,分析对2种损失函数的基于支持向量机的波束形成算法.仿真实验结果表明,该算法在期望信号方向存在指向误差时,依然能够保持较好的系统输出信噪比,具有一定的稳健性.
A robust beamforming method is proposed,which can effectively overcome the influence of DOA(direction of arrival)mismatch.Based on the analysis of linearly constrained minimum variance beamformer,loss function is introduced in optimization problem of linearly constrained minimum variance beamformer.Then the support vector regression(SVR)algorithm is applied to the robust beamforming,which is based on the principle of structural risk minimization.Under an ideal scenario of no-mismatch and an actual scenario of mismatch respectively,the SVR-based beamforming method of two loss function are researched via Matlab simulation through choosing the different parameters and signal to noise ratio(SNR).Simulation results show that the SVR-based beamforming method enhances the robustness in terms of desired signal array manifold vector errors.
出处
《西安工程大学学报》
CAS
2014年第5期598-603,共6页
Journal of Xi’an Polytechnic University
基金
西安工程大学博士科研启动基金(BS1414)
关键词
波束形成
线性约束最小方差
支持向量机
beamforming
linearly constrained minimum variance(LCMV)
support vector regression(SVR)