基于遗传优化的正交小波分数间隔盲均衡算法

摘要

Aimed at some disadvantages of traditional constant modulus blind equalization algorithm (CMA), such as low convergence speed, large steady-state error and local convergence, an orthogonal wavelet transform fractionally spaced blind equalization algorithm based on genetic optimization algorithm (WT-FSE-GA) is proposed. It can reduce the signal autocorrelation by orthogonal wavelet transform to input signal of fractionally spaced blind equalizer. And, it can also reduce the possibility of CMA's local convergence using characteristics of random global search of genetic algorithm to optimize the equalizer weight vector. It has faster convergence speed and smaller mean square error compared with FSE-CMA and WT-FSE. Its efficiency is proved by using numerical simulation.%针对传统的常数模盲均衡算法(CMA)收敛速度慢、稳态误差大和局部收敛的缺点,提出了一种基于遗传优化的正交小波分数间隔盲均衡算法(WT-FSE-GA).该算法通过对分数间隔均衡器的输入信号进行正交小波变换,以降低信号的自相关性,并利用遗传算法全局随机搜索的特性,对均衡器权向量进行优化,以降低CMA陷入局部收敛的可能性.与分数间隔盲均衡算法(FSE-CMA)、正交小波变换分数间隔盲均衡算法(WT-FSE)相比,该算法收敛速度更快、稳态误差更小.水声信道的仿真结果,验证了该算法的有效性.