用于信道盲均衡的基于软判决导向的LS算法

本文提出两种改进的基于软判决导向的最小平方均衡法。一种采用横向结构，另一种采用格式结构。这两种均衡算法的判决误差形式不同于一般的判决误差形式，它们的表达式是一种更为复杂的连续函数形式。通过计算机仿真，我们发现，采用横向结构的这种改进算法无论对于最小相位还是非最小相位信道的均衡均有较好的性能，而采用格形结构的这种改进算法的性能在非最小相位信道下的性能反而不如传统的格形算法。

DECISION DIRECTED ALGORITHMSFOR BLIND EQUALIZATION

In this paper,we present two new LS (Least Square) algorithms based on the idea of soft decision We modified the classical LS FTF(Least Square Fast Transversal Filter)and LSLL(Least Square Lattice Ladder)algorithm by substituting the ordinary decision error function into a more complex and general decision error expression We call the modified LS algorithm as the soft decision directed LS algorithm We use it in the case of blind equalization Through computer simulations,we find that the Soft Decision Directed LS FTF algorithm exhibits a better performance than the ordinary FTF algorithm in both minimum and nonminimum phase channels whereas the Soft Decision Directed LSLL algorithm does not make a better performance than the classical LSLL algorithm in non minimum phase channel