时变步长自适应均衡算法的设计与仿真

研究了自适应均衡算法中步长因子对算法性能的影响。步长因子大，算法收敛速度和跟踪速度快，但收敛后稳态剩余误差大；步长小，算法收敛速度和跟踪速度慢，收敛后稳态剩余误差小。因此，采用固定步长时，算法在收敛速度和收敛精度方面对步长的要求是相互矛盾的。因此本文采用均方误差的Sigmoid函数和误差信号的非线性函数作为步长控制因子，得到了几种改进算法，经计算机仿真表明，改进算法收敛速度加快，稳态剩余误差减小。

Design and Simulation of Variable Step-size Adaptive Equalization Algorithm

The effect of step-size factor on the performance of adaptive equalization algorithm is researched in the paper.If the step-size factor is big,convergence rate and track rate of the algorithm are faster,but state residual error is bigger after convergence.By contrast,if the step-size factor is small,convergence rate and track rate of the algorithm is slow,but state residual error is smaller after convergence.Therefore,the demand for step-size is contrary between convergence rate and convergence precision if the fix step-size is adopted.The Sigmoid function of mean square error and non-linear function of error signal are adopted as step-size control factor,and several algorithms are got.Simulation results show that convergence rate of the improved algorithm is quickened and state residual error is diminished.