广义特征值盲波束形成的低复杂度自适应算法

提出了一种线性低复杂度盲自适应Lagrange乘子波束形成算法．基于最大信干燥比准则(MSINR)的特征值波束形成将导致广义特征值(GE)问题，新算法通过把GE问题表示成期望信号和干扰噪声信号协方差矩阵特征值的函数，用线性迭代的方法搜索最大GE，并利用最陡下降法更新权向量；提出方法被用于W—CDMA智能天线基站上行信号接收．数值仿真结果表明，与其他算法相比，在未牺牲系统BER性能的同时，提出算法具有更快的收敛速度和更低的计算复杂度；算法总的计算复杂度约为O(7N)(N为天线元数目)．

Low Complexity Adaptive Algorithm of the Generalized Eigenvalue Blind Beamforming

A blind adaptive Lagrange multiplier beamforming algorithm was proposed, which solves generalized eigenvalue(GE) problem with linear and low computational load. Eigen-beamforming based on maximizing the signal to interference plus noise ratio(MSINR)criteria results in the GE problem. After the objective GE problem is expressed by the eigenvalues of the covariance matrixes of the desired signal and interfering plus noise signal vector, a linear and iterative adaptive procedure is developed to search the maximal generalized eigenvalue, and the weight vector is updated by the steepest descent method. The proposed algorithm was applied to the base station of W-CDMA smart antenna system for uplink receiving. The simulation results show that, without sacrificing the bit error rate (BER) performance of the array system, the proposed algorithm can achieve more fast convergence rate and lower computational complexity than other beamforming algorithms, and the total computational load is about O(7N) where N is the number of antenna elements.