Design & Implementation of Feed Forward Neural Network for Fir Filter von Waseem, Mohammed: New (2013)

Beschreibung

pp. 72.

Bestandsnummer des Verk�ufers 26128813638

Verk�ufer kontaktieren

Diesen Artikel melden

Bibliografische Details

Titel: Design & Implementation of Feed Forward Neural Network for Fir Filter
Verlag: VDM Verlag Dr. Mueller Aktiengesellschaft & Co. KG
Erscheinungsjahr: 2013
Sprache: Englisch
ISBN-10: 3659449237
ISBN-13: 9783659449239
Einband: Softcover
Zustand: New

�ber diesen Titel

Inhaltsangabe

Filter designs have number of applications in data transmission systems, perfect reconstruction filter banks, nonuniform sampling, interpolation filters over the past two decades. There are two conventional methods, which are FIR and IIR filter forms, to design filters. FIR filters can be designed with an exact linear phase. However, when the sharp magnitude specifications are required, higher order FIR filters are generally needed, and a larger delay results. On the other hand, IIR filters have two disadvantages: one is the stability that must be considered, and another is that the existing design methods are generally time consuming. For real-time signal applications, the above methods are all linear algebra based methods, therefore, cannot meet the requirements of real-time. Neural networks possessing parallel processing capability have been successfully applied for solving various computationally expensive optimization problems. This is due to the properties of guaranteed convergence to a local minimum of the Lyapunov energy function and the fast computational speed when implemented in hardware.

Rese�a del editor

Filter designs have number of applications in data transmission systems, perfect reconstruction filter banks, nonuniform sampling, interpolation filters over the past two decades. There are two conventional methods, which are FIR and IIR filter forms, to design filters. FIR filters can be designed with an exact linear phase. However, when the sharp magnitude specifications are required, higher order FIR filters are generally needed, and a larger delay results. On the other hand, IIR filters have two disadvantages: one is the stability that must be considered, and another is that the existing design methods are generally time consuming. For real-time signal applications, the above methods are all linear algebra based methods, therefore, cannot meet the requirements of real-time. Neural networks possessing parallel processing capability have been successfully applied for solving various computationally expensive optimization problems. This is due to the properties of guaranteed convergence to a local minimum of the Lyapunov energy function and the fast computational speed when implemented in hardware.

��ber diesen Titel� kann sich auf eine andere Ausgabe dieses Titels beziehen.