摘要
在投影波束形成的基础上,提出了一种基于主成分分析神经网络的波束形成算法。该方法利用自适应迭代形式获得了波束形成的权向量,避开了投影算法对相关阵特征值分解的计算过程,因而减少了计算量,获得了相对快速的波束形成图,提高了自适应能力。经数值模拟仿真,验证了该方法的正确性。
In this paper, a kind of algorithm of beam-forming for neural network is presented based on the principal component analysis on the basis of projection beam-forming model. The beam-forming weight vectors are gained by the self-adaptive iteration method, and avoiding the need for the process of eigen-value decomposition. Therefore, it reduces the computational load and obtains not only the relatively faster beam-forming diagram but also better adaptive performance. The computer simulations demonstrate the effectiveness of the proposed method.
出处
《山东科技大学学报（自然科学版）》
CAS
2008年第5期57-60,共4页
Journal of Shandong University of Science and Technology(Natural Science)
基金
山东省自然科学基金项目(Y2006F36)
关键词
主成分分析
神经网络
波束形成
principal component analysis
neural network
beam-forming
