In the direct blind MMSE equalization algorithm based on cross-correlation vector estimation, two auto-correlation matrix of the received signal is usually used to compute the outer-product of cross-correlation vector. A novel blind MMSE equaliza tion algorithm is proposed, which requires only one auto-correlation matrix to obtain the outer-product based on the property of shift matrix. The computational complexity is reduced by two times. Then an efficient fast adaptive implementation of the novel algorithm is presented to recursively update the equalizer coefficient vector real time. It has the computational complexity of O( K2) . Numeri cal simulations show that the proposed algorithm has better performance then the old algorithm and it is robust to channel order mis match. Furthermore, the adaptive implementation has good convergence performance.%基于互相关向量估计的直接盲MMSE均衡算法通常需要观测向量两个相邻延时的自相关矩阵来构建互相关向量外积.本文提出了一种盲MMSE均衡新算法,利用了统计独立输入信号自相关矩阵是移位矩阵的特殊性质,只需观测向量单个延时的自相关函数便可得到互相关向量的外积,使计算量减少二分之一.然后推导出了新算法的自适应实现过程,实时地对均衡系数向量更新估计,每次更新计算量只有O(K2).仿真证明新算法的性能优于原算法,而且对信道阶数错误估计非常稳健,自适应实现算法也具得很好的收敛性能.
展开▼
