基于神经网络的波束形成

Neural Network-based Beamforming

本文给出利用神经网络来实现智能天线的波束形成的方法。选取了适当的神经元数目 ,减少了训练时间。模拟输出的赋形结果与维纳 (Wiener)解相吻合 ,并与传统算法进行性能比较 ,表明了该方案的有效性。

This paper proposes a approach of using neural network for smart antennas. Appropriate neural nets are used in order to reduce the network training time. The results of simulating beamforming are agree well with the Wiener solution , performance of this scheme is compared with conventional algorithm, and confirming the validity of this method.