基于标准支持向量回归的阵列波束优化研究

Research on Optimization of Array Beamforming Based on Standard Support Vector Regression

针对支持向量机在解决小样本、非线性和高维模式处理问题上的优势,将支持向量回归与波束优化理论进行对比,修正支持向量回归价值损失函数,分析支持向量回归波束优化的基本方法及其应用条件,建立标准支持向量回归波束优化模型,研究了基于标准支持向量回归波束形成器的优化模型及具体实现过程,并进行了数值仿真实验.仿真实验结果表明,在不同的阵型、不同的价值损失函数、不同的数据样本和不同的信噪比下,基于标准支持向量回归的波束形成器在指向性和旁瓣级等性能指标上均取得了较好的效果,为波束形成器的优化设计提供了一种有效且可行的方法.

For the advantages of support vector machines,such as the small sample,nonlinear and high dimensions, the paper compares the support vector regression algorithm with beam optimization theory ,modifies the cost function of support vector regression,and analyzes the basic methods and its application conditions to optimizing the beam of support vector regression. Then the optimization model of beamformer based on standard support vector regression is established,the optimization model and the concrete implementation process of beamformer based on the standard support vector regression is discussed,and the numerical simulation experiments are done. The results show that for different arrays,different cost functions,different data samples and different SNR,the beamformer based on standard support vector regression has achieved good results on many performances,such as directivity,sidelobe level. Compared with the conventional beamforming,it provides an effective and feasible method for the optimization design of beamformer.