Title :
Dynamic Convergence of LMS Algorithm Using New Step Size
Author :
Gurung, J.B. ; Khargekar, A.K. ; Kumaravelu, P.G.
Author_Institution :
KNS Inst. of Technol., Bangalore
Abstract :
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 the LMS algorithm is also minimized to the large extends.
Keywords :
convergence of numerical methods; least mean squares methods; LMS algorithm; dynamic convergence; dynamically varying parameter; new step size; Adaptive signal processing; Autocorrelation; Convergence; Equalizers; Iterative algorithms; Least squares approximation; Resonance light scattering; Signal processing; Signal processing algorithms; Stability;
Conference_Titel :
Integrated Circuits, 2007. ISIC '07. International Symposium on
Conference_Location :
Singapore
Print_ISBN :
978-1-4244-0797-2
Electronic_ISBN :
978-1-4244-0797-2
DOI :
10.1109/ISICIR.2007.4441822