双模式变步长解相关盲均衡算法研究

恒模算法(CMA)是一种重要的盲均衡技术,虽然能达到盲均衡效果,但是收敛后会有较大的剩余误差和相位模糊。在文献3]中提出的VS-MCMA+DD-LMS算法的基础上做了进一步改进,提出变步长修正多模算法(VS-MMMA)和DD-LMS相结合,以及变步长解相关修正多模算法(VS-UMMMA)和DD-LMS相结合2种新算法,进行了仿真和性能分析,证明新算法在收敛速度和均衡误差方面均有明显的改进。

Research of a Variable Step Dual-mode Decorrelating Blind Equalization Algorithm

CMA is a very important technology of blind equalization.Although it can achieve the aim of blind equalization,it has the problems of large residual error and phase ambiguity when converged.In this paper,two variable step modified dual-mode blind equalization algorithms are proposed.The algorithm of VS-MMMA＋DD-LMS combines the modified multi-modulus algorithm and the decision directed algorithm.The algorithm of VS-UMMMA＋DD-LMS combines the decorrelating modified multi-modulus algorithm and the decision directed algorithm.The algorithms proposed in this paper can improve the speed of convergence and reduce the residual error effectively compared with the algorithm（VS-MCMA＋DD-LMS） in paper 3].The capability of improved algorithms are simulated and analysed in this paper.