Novel steepest descent adaptive filter derived from new performance function with additional exponential term
نویسندگان
چکیده
A new steepest descent adaptive filter algorithm derived from a newly devised performance index function is presented. The performance function of the new algorithm is introduced from that of the least mean square (LMS) considering that the stochastic steepest descent method utilises a gradient search in order to minimise the performance function iteratively. Through mathematical analyses and computer simulations, it is verified that there are significant improvements in convergence speed and misadjustment error. Nevertheless its computational simplicity and robustness are maintained with little degradation (compared to those of the LMS algorithm). The new algorithm can be interpreted as a new kind of variable step size adaptive algorithm, and in this respect a modified method is proposed in order to reduce the noise caused by fluctuation of the varying step size. Zusammenfassung In diesem Artikel wird ein neuer Steepest-Descent-Algorithmus f/Jr adaptive Filter vorgestellt, der sich aus einer neu konstruierten Leistungsindexfunktion ableiten l/iBt. Die Leistungsfunktion des neuen Algorithmus wird fiber jene des LMS eingefiJhrt, wobei berficksichtigt wird, dab die stochastische Steepest-Descent-Methode eine Gradientensuche ben/izt, um die Leistungsfunktion schrittweise zu minimieren. Mittels mathematischer Analysen und Computersimulationen wird nachgewiesen, dab deutliche Verbesserungen hinisichtlich der Konvergenzgeschwindigkeit und des Misadjustments auftreten. Die rechnerische Einfachheit und Robustheit bleibt mit einer nur geringen Verschlechterung gegenfiber jener des LMS-Algorithmus dennoch erhalten. Der neue Algorithmus kann als eine neue Art eines adaptiven Algorithmus mit ver/inderlicher Schrittweite interpretiert werden, und in diesem Zusammenhang wird eine modifizierte Methode vorgeschlagen, um das durch die Schwankungen der ver/inderlichen Schrittweite verursachte Rauschen zu verringern. R~um~ Un nouveau algorithme de filtrage adaptatif h descente rapide d6riv6 d'une nouvelle fonction ~i index de performance est pr6sent6. La fonction de performance du nouvel algorithme est introduite ~i partir de la LMS en consid6rant que la m6thode de descente rapide stochastique utilise une recherche de gradient dans le but de minimiser it6rativement la *Corresponding author. Tel: + 82-42-869-3438, Fax: + 82-42-869-3410. 0165-1684/94/$7.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0165-1684(93)E0085-Y 190 B.-E. Jun, D.-J. Park / Signal Processing 36 (1994) 189-199 fonction de performance. Au travers d'analyses mathrmatique et de simulations, il est vrrifi6 qu'il y a des amrliorations significatives dans la vitesse de convergence et dans l'erreur d'ajustement. Nranmoins, sa simplicit6 de calcul et sa robustesse sont maintenues avec peu de drgradation fi partir de l'algorithme LMS. Le nouvel algorithme peut 6tre interprrt6 comme un nouveau type d'algorithme adaptatif fi pas variable, et de ce fait, une mrthode modifire est proposre dans le but de rrduire le bruit caus6 par la fluctuation de la taille du pas variable.
منابع مشابه
Block Lms Adaptive Filter with Deterministic Reference Inputs for Event-related Signals
Adaptive estimation of the linear coefficient vector in truncated expansions is considered for the purpose of modeling noisy, recurrent signals. The block LMS (BLMS) algorithm, being the solution of the steepest descent strategy for minimizing the mean square error in a complete signal occurrence, is shown to be steady-state unbiased and with a lower variance than the LMS algorithm. It is demon...
متن کاملMultiple environment optimal update profiling for steepest descent algorithms
In this paper, the methods for use of prior information about multiple operating environments, in improving adaptive filter convergence properties are discussed. More concretely, the gain selection, profiling and scheduling in steepest descent algorithms are treated in detail. Work presented in this paper is an extension of [1]. Two flavors of optimization are discussed: average descent rate op...
متن کاملThe least mean fourth (LMF) adaptive algorithm and its family
New steepest descent algorithms for adaptive filtering and have been devised which allow error minimization in the mean fourth and mean sixth, etc., sense. During adaptation, the weights undergo exponential relaxation toward their optimal solutions. T ime constants have been derived, and surprisingly they turn out to be proportional to the time constants that would have been obtained if the ste...
متن کاملAdaptive algorithms for Weighted Myriad Filter optimization
Stochastic gradient-based adaptive algorithms are developed for the optimization of Weighted Myriad Filters, a class of nonlinear lters, motivated by the properties of -stable distributions, that have been proposed for robust non-Gaussian signal processing in impulsive noise environments. An implicit formulation of the lter output is used to derive an expression for the gradient of the mean abs...
متن کاملAdaptive Weighted Myriad Filter Algorithms forRobust Signal Processing in - Stable
Stochastic gradient-based adaptive algorithms are developed for the optimization of Weighted Myriad Filters. Weighted Myriad Filters form a class of nonlinear lters, motivated by the properties of-stable distributions, that have been proposed for robust non-Gaussian signal processing in impulsive noise environments. The weighted myriad for an N-long data window is described by a set of non-nega...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- Signal Processing
دوره 36 شماره
صفحات -
تاریخ انتشار 1994