Open AccessA Deep Convolutional Learning Method for Blind Recognition of Channel Codes
Author(s) -
Feng Tian,
Jiao Wang,
Jianqing Li
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 - 10.1088/1742-6596/1621/1/012088
Subject(s) - computer science , convolutional neural network , robustness (evolution) , artificial intelligence , deep learning , convolutional code , channel (broadcasting) , pattern recognition (psychology) , speech recognition , machine learning , algorithm , decoding methods , telecommunications , biochemistry , chemistry , gene
Blind identification of channel codes is a significant part in the field of non-cooperative signal processing. It plays a significant role in intelligent communication, information interception, information confrontation and so on. To solve issues of high dependence of manual extraction of expert features, low robustness, difficulty of deployment from traditional methods for blind recognition of channel codes, this paper presents a novel way for blind recognition of channel codes by integrating deep learning technology. For the extraction of features from signals, deep convolutional neural network containing 4 convolutional layers is designed. For model training, database of convolutional codes is generated. Experimental results show that this method can achieve 98%+ recognition accuracy with SNR not lower than 4 dB, which proves the effectiveness of the method.