DocumentCode :
3545162
Title :
A new architecture for analog sampled-data neural filters
Author :
Sedighi, Behnam ; Analui, Behnam ; Sharif-Bakhtiar, Mehrdad
Author_Institution :
Sharif Univ. of Technol., Tehran, Iran
fYear :
2005
fDate :
23-26 May 2005
Firstpage :
2494
Abstract :
A new architecture for analog sampled-data filters is presented. This architecture is based on implementing the difference equation of the filter using a neural network. It results in lower sensitivity of filter coefficients to hardware nonidealities. Different tradeoffs in the hardware implementation of the filter are discussed. A design example is also presented.
Keywords :
Butterworth filters; FIR filters; IIR filters; band-pass filters; neural nets; sampled data filters; switched capacitor filters; FIR filter; IIR sampled-data filter; analog sampled-data neural filters; filter difference equation; linear neural filters; neural networks; second-order band-pass Butterworth filter; switched-capacitor filters; switched-current filters; Circuit optimization; Difference equations; Dynamic range; Electrical engineering; Frequency domain analysis; Hardware; IIR filters; Neural networks; Nonlinear filters; Transfer functions;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Circuits and Systems, 2005. ISCAS 2005. IEEE International Symposium on
Print_ISBN :
0-7803-8834-8
Type :
conf
DOI :
10.1109/ISCAS.2005.1465132
Filename :
1465132
Link To Document :
بازگشت