一种基于模糊神经网络的盲均衡算法

提出一种新的基于模糊神经网络的盲均衡算法,它利用盲估计算法与模糊神经网络相结合的原理,先对通信信道进行盲估计,利用得到的信道特性来估计FNN盲均衡器中的隶属函数中心与宽度初值。传统算法中,中心与宽度的初值设定由专家提供,会有一定的偏差。本算法利用信道特性来估计隶属函数的初值,有效地避免了人为确定初值的不确定性。同时本算法克服了原有盲均衡算法反卷积计算量大的缺点。仿真结果证明本算法与同类算法相比,收敛速度加快,误码率减小。

A Blind Equalization Algorithm Based on the Fuzzy Neuron Network

In this paper,a new blind equalization algorithm based on fuzzy neural network(FNN)is proposed.It makes use of blind estimation(BE)and FNN to equalize.Firstly BE algorithm is used to identify the channel character,which is used to estimate the expectation and the variance.In traditional algorithm,the initial values of the expectation and the variance are provided by expert,which can lead to decide signals incorrectly.The algorithm estimate the initial value using the channel character,so it can avoid indetermination and ensure to decide correctly.At the same time,this algorithm overcomes the defect of great computation of inverse convolver.Results of the simulation show that the algorithm has faster convergence speed and lower bit error rate(BER)than the other algorithms.