基于周期平稳信号特性的神经网络波束形成技术  

Neural Network Beamforming Technique for Cyclostationary Signals

作　　者：陈宇欣[1] 何振亚[1] 

机构地区：[1]东南大学

出　　处：《应用科学学报》1998年第4期403-408,共6页Journal of Applied Sciences

基　　金：国家自然科学基金;国家攀登计划重大关键资助

摘　　要：绝大多数通信信号都具有周期平稳信号特性，ＳＣＯＲＥ算法就是一种利用周期平稳信号特性的优良的波束形成算法．然而，对于矩阵运算和矩阵求逆的巨大运算量的要求限制了它的应用．文中根据神经网络理论提出了一种神经网络波束形成技术（ＮＮＢＦＴ），较好地解决了这一问题．仿真结果表明其性能优良。Most of the communication signals have the cyclostationary character. SCORE is an excellent algorithm based on cyclostationary. However, main disadvantage of this algorithm is the huge calculation of the matrix compution and the matrix inverse compution. A neural network beamforming technique which can solve this problem successfully is proposed in this paper. The experiment results have shown that this algorithm has better performance and can reject noise strongly.

关 键 词：波束形成 神经网络 周期平稳信号 信号处理 

分 类 号：TN911.7[电子电信—通信与信息系统]