In the MCMA, adjusting the scale coefficients of two gradient vectors can improve the convergence rate and reduce residualmeansquare error.
调整该算法中两个梯度矢量的比例系数,可提高该算法收敛速度、减少收敛后的均方误差。
2
Though the decision directed least meansquare error algorithm (DD-LMS) has small residual error, when the channel eye close, it can not converge. That is, it has no cold start ability.