基于NCMA算法的递归步长多模盲均衡算法

为了解决传统常数模盲均衡算法收敛速度缓慢的问题，考虑到M-QAM信号模值分布在几个已知半径圆上的特点，该文给出了一种新的基于归一化常数模（NCMA）的递归步长多模盲均衡算法。这种算法能够自适应地调节步长，使收敛后的步长在达到最优的同时，得到的剩余稳态误差也达到最小。仿真实验表明，该算法能够在计算量增加不大的前提下，具有较好的收敛性能及抗误码性能，其收敛速度也得到较大的改善。

A Recursive Step Multi-Modulus Algorithm for Blind Equalization Based on Normalized CMA

Considering the characteristic of the M-QAM signals which lies on several circles with known radius,this paper presents a new blind equalization algorithm which combines the benefits of the Normalized constant modulus algorithm(NCMA) and the recursive step multi-modulus algorithm.In respect that this algorithm can adjust step automatically,the ultimate step can reach an optimum value and remnant error can arrive at minimum value simultaneity.In addition,in order to speed up the convergence process,this algorithm can switch over to a Decision-Directed equalization scheme once the error level is reasonably low.The computer simulation results demonstrate fast convergence and satisfactory performance with these proposed algorithms.