模糊神经网络分类器在盲均衡算法中的应用  

Application of fuzzy neural network classifier in blind equalization algorithm.

作　　者：孙云山[1] 李艳琴[2] 张立毅[1] 

机构地区：[1]天津商业大学信息工程学院,天津300134 [2]天津大学电子信息工程学院,天津300003

出　　处：《计算机工程与应用》2008年第7期171-173,共3页Computer Engineering and Applications

基　　金：中国博士后基金(the Post-doctor Foundation of China under Grant No.20060390170);天津市高等学校科技发展基金(the Tianjin College Science and Technology Development Foundation of China under Grant No.20060610)

摘　　要：提出一种基于模糊神经网络分类器的盲均衡算法,将盲信道估计与模糊神经网络分类器相结合,先对通信信道进行盲估计,然后利用卷积原理重建信号,用模糊神经网络替代原有的判决器,从而实现了盲均衡。通过仿真实验证明,该算法加快了收敛速度,减小了剩余误差,降低了误码率。A blind equalization algorithm based on fuzzy neural network is proposed.Blind channel estimation and fuzzy neural network classifier are utilized to realize blind equalization.Firstly blind channel estimation is used to identify the character of the channel.Signals are rebuilt by de-convolution,and the original judgment equipment is replaced by fuzzy neural network classifier. Simulations indicate that the novel algorithm improves convergence and bit error rate and so on.

关 键 词：盲均衡 模糊神经网络 信道估计 分类器 

分 类 号：TN914.5[电子电信—通信与信息系统]