基于归一化常数模算法的级联自适应盲均衡算法

为了克服常数模算法收敛速度慢.稳态误差大的缺点,利用分数间隔盲均衡器的优点,分析了过采样的分数间隔盲均衡器理论,提出了一种基于归一化常数模算法的级联自适应盲均衡算法.该算法以分数间隔盲均衡器为第一级,以波特间隔盲均衡器为第二级,由归一化常数模算法对两级均衡器权向量进行更新,用水声信道对算法的性能进行了仿真研究.结果表明,分数间隔盲均衡器的性能优于波特间隔盲均衡器;高采样率的分数间隔盲均衡器与波特间隔盲均衡器级联后的性能优于低采样率的分数间隔盲均衡器与波特间隔盲均衡器级联后的性能,而且级联盲均衡器总是优于单级的分数间隔盲均衡器.

Cascaded Adaptive Blind Equalizer Based on Normalized Constant Modulus Algorithm

To overcome the low convergent rate and high mean square error of the Baud Spaced Equalizer based on Constant Modulus Algorithm (BSE-CMA),on the basis of the advantage of Fractionally Spaced Equalizer based on Constant Modulus Algorithm (BSE-CMA),the principle of over-sampling fractionally spaced equalizer was analyzed,and Cascaded adaptive Blind Equalizer based on Normalized Constant Modulus Algorithm (CBE-NCMA) was proposed. In the proposed CBE-NCMA,the Fractionally Spaced Equalizer (FSE) was used as its first level whereas Baud Spaced Equalizer (BSE) was employed for its second level,and the weight vectors of these two equalizers were updated using the Normalized Constant Modulus Algorithm(NCMA). The performance of the CBE-NCMA was simulated using underwater acoustic channel. The results show that the FSE-NCMA outperforms the BSE-NCMA in compensating the distorted channel,that CBE-NCMA exceeds FSE-NCMA,and that the higher the sampling rate is,the better the CBE-NCMA is.