基于双曲正切误差函数的变步长盲均衡算法

Hyperbolic tangent error function based variable step size blind equalization algorithm

为了克服常数模算法(CMA)收敛速度慢,收敛后均方误差大的不足,定义了双曲正切误差函数,提出了基于双曲正切误差函数的变步长盲均衡新算法。该算法利用双曲正切误差函数关于均衡器输出幅度等于常数模的轴奇对称特性来减少均方误差,利用变步长来加快收敛速度。用负声速梯度水声信道,对算法的性能进行了仿真研究。结果表明,该算法收敛速度快、均方误差小,较好地克服了CMA的缺陷。

For greatly reducing the MSE （Mean Square Error）of CMA（ Constant Modulus Algorithm） at the end of the convergence, HE （ Hyperbolic tangent Error function） is defined, HEVBE （ Hyperbolic tangent Error function based Variable step size Blind Equalization algorithm） proposed. The HEVBE has fast convergent speed by variable step size, and the small MSE duo to odd symmetric feature of the HE about axis that output modulus of equalizer is equal to constant modulus, The efficiency of the HEVBE is proved by computer simulation of negative acoustic gradient in underwater acoustic channels.