Presentation 1998/12/10
Simplification of Neural Network Equalizer Using Genetic Algorithm
Hisashi Osawa, Tatsuro Kawamoto, Yoshihiro Okamoto, Hidetoshi Saito,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A searching method using genetic algorithm is studied for simplification of neural network equalizer. First. a model of read/write channel with partial erasure and a searching method using genetic algorithm are described. Then, the relationship between the required SNR to achieve a bit error rate of 10^<-4> and the reduction parameter for the 16/17(0.6/6)coded EPR4ML system equalized by a simplified neural network is obtained by computer simulation and is compared with that of the conventional neural network equalizer. The results show that the performance of simplified neural network equalizer is equal to that of conventional neural network equalizer.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) digital magnetic recording / neural network / genetic algorithm / partial erasure / PRML
Paper # MR98-37
Date of Issue

Conference Information
Committee MR
Conference Date 1998/12/10(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Magnetic Recording (MR)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Simplification of Neural Network Equalizer Using Genetic Algorithm
Sub Title (in English)
Keyword(1) digital magnetic recording
Keyword(2) neural network
Keyword(3) genetic algorithm
Keyword(4) partial erasure
Keyword(5) PRML
1st Author's Name Hisashi Osawa
1st Author's Affiliation Faculty of Engineering, Ehime University()
2nd Author's Name Tatsuro Kawamoto
2nd Author's Affiliation Faculty of Engineering, Ehime University
3rd Author's Name Yoshihiro Okamoto
3rd Author's Affiliation Faculty of Engineering, Ehime University
4th Author's Name Hidetoshi Saito
4th Author's Affiliation Faculty of Engineering, Ehime University
Date 1998/12/10
Paper # MR98-37
Volume (vol) vol.98
Number (no) 454
Page pp.pp.-
#Pages 8
Date of Issue