摘要
利用径向基函数神经网络来实现无线通信中智能天线的波束形成。利用了信号相关阵的对称性质 ,仅考虑相关阵中的上三角 (或下三角 )的部分元素作为网络的输入量 ,计算量较小。进一步提高了运算速度 ,具有收敛快且准确的优点。计算机模拟实例证明结果与理论相吻合 ,表明了该方案的有效性。
This paper proposes an approach of the blind beamforming with the radial basis function neural network for smart antennas in the wireless mobile communications. The symmetry property in the correlation matrix is exploited to reduce the dimension of the input vectors.One needs only either the upper (or the lower triangular) half of the correlation matrix. The advantages of the proposed method include reductions in computation with fast converging speed and the quality of being precise .The simulating results are agree well with the theory and confirm the validity of this method.
出处
《电波科学学报》
EI
CSCD
2004年第2期237-239,252,共4页
Chinese Journal of Radio Science
