Open AccessNeural net decoders of nonbinary codes
Author(s) -
V V Butov,
V. N. Dumachev,
S E Fedyaeva
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1479/1/012089
Subject(s) - classifier (uml) , artificial neural network , computer science , feature vector , algorithm , error detection and correction , parity (physics) , pattern recognition (psychology) , mathematics , artificial intelligence , physics , particle physics
In paper, neural net decoders of nonbinary error correction codes are considered. Analytic methods for calculating of synapse weight coefficients are proposed. It is shown that for codes ( n , k ) with a small number of corrected errors ( n − k ≪ k ), it is advisable to use a 6-layer universal classifier based on the feature space of parity symbols. For codes with k ≪ n − k , it is proposed to use a 3-layer classifier on feature space of information symbols.