基于不同奇对称误差估计的盲均衡算法

为了研究盲均衡算法中误差函数形式对盲均衡器性能的影响,定义并分析了不同奇对称误差函数的特点,提出了基于不同奇对称误差估计的变步长盲均衡新算法（DOSVCMA）。该算法充分利用误差函数的对称性来减小均方误差,利用变步长来加快收敛速度,在步长特性参数相同的前提下,用深海声道轴信道和双径水声信道对该算法进行了仿真研究。结果表明,用不同奇对称误差函数设计的均衡器,在收敛速度和收敛后均方误差方面的性能是不同的,但比常规的变步长盲均衡算法（VCMA）的性能优越,这一研究为提高盲均衡器性能提供了一条途径。

Blind Equalization Algorithm with Variable Step Size Based on Different Odd Symmetry Error Estimate

For studying the influence of different error functions on the performance of blind equalization,different odd symmetry error functions were defined and their features were analyzed,and DOSVCMA(Different Odd Symmetry error estimate based Variable step size Constant Modulus blind equalization Algorithm) was proposed. In DOSVCMA,the mean square error(MSE) could be greatly reduced by using the symmetry of error functions and the convergent speed could be improved based on the performance of variable step size. Under condition of the same parameters charactering variable step size,simulation tests with axial channel with deepwater acoustic track and double path underwater acoustic channel were carried out. The results show that the equalizers designed by different odd symmetry error functions have different performances in convergence speed and MSE and the DOSVCMA outperforms the variable step size based traditional CMA(Constant Modulus Algorithm) in reducing the MSE and improving the convergent speed. Accordingly,the research results provide a means for improving the performance of blind equalizer.