A new variable step-size transform domain LMS algorithm with system identification

This paper presents a new variable step-size transform domain LMS (NVSTDLMS) algorithm. In the algorithm, a new variable step size is timely adjusted by employing all the output errors available so far. Therefore, the memories of the previous learning information are not lost and fully used to adjust the current value of the step size. What is more, the ratio form of the computation with two different convergence factors reduces the influences of the additive noise on the step-size adjustment. As a consequence, the new algorithm has favorable performance. Simulation results for system identification are presented to demonstrate the performance improvements in convergence speed and steady-state misadjustment compared with other existing approaches.