Dynamic Convergence of LMS Algorithm Using New Step Size

LMS Algorithm already has a standard step-size, mu. This paper still optimizes the existing step-size, while it is under processing. It simply introduces a dynamically varying parameter at the denominator of the step-size so that the convergence rate of the system will drastically improve. Mean time this empirical result shows that the unstable problem of… (More)