摘要
本文通过将约束和非约束频域分组LMS算法的权值迭代方程转换成时域方程,利用最小二乘法(LS)分别选取最优时变步长收敛因子,得到频域最佳分组算法(FOBA)和非约束频域最佳分组算法(UFOBA),虽然增加了计算量,但计算机仿真结果表明:提高了收敛速度和精度,算法稳定可靠。
In this paper, the updating equations of FBLMS and UFBLMS algorithms are transformed into the time domain, then two algorithms(FOBA and UFOBA) are presented by employing time-varying convergence factors which are optimized in a least-square(LS) sense respectively. Although FOBA and UFOBA algorithms require a relatively modest increase in computation for each block iteration compared to FBLMS and UFBLMS algorithms respectively, it is shown by simulations that these-algorithms improve convergence speed and accucary of adaptation.
关键词
自适应信号处理
LMS算法
收敛因子
Adaptive signal processing, LMS algorithm, Convergence factor
