BCH code decoding method based on deep learning

Abstract

The invention relates to a BCH code decoding method based on deep learning. The method comprises the following steps: receiving and extracting quantization level information from a channel of a BCH code to serve as reliability information, then performing hard decision on a code word, calculating the obtained code word information by using a check matrix to obtain a syndrome of the BCH code, and converting the syndrome into a binary form; gradually inputting binary syndrome information into two bidirectional GRUs connected in series of the neural network at one input end in a time step manner;at the other input end, inputting bit-level quantization level information as reliability into the auxiliary syndrome information for decoding; merging the information output by the MLP and the reliability into new two-dimensional matrix information, inputting the new two-dimensional matrix information into another GRU in the next step to enable the neural network to find part of error positionsfrom two influence factors of reliability and a syndrome, and connecting a full connection layer to the output end of the GRU to flatten the data; and obtaining neural network decoder.

Description

technical field [0001] The invention belongs to the field of error control coding in channel coding, and relates to a deep learning decoding algorithm using BCH code syndrome and level reliability information. Background technique [0002] With the continuous development of science and technology and the continuous progress of society, people put forward higher and higher requirements for the reliability of communication system information transmission. In 1948, American mathematician Shannon proposed the concept of information entropy in his article "Mathematical Theory in Communication", which laid a theoretical foundation for information encoding. In the next few years, Hamming code, Golay code and other coding methods appeared in channel coding, and achieved fruitful results. In 1959 and 1960, Hocquenghem, Bose and Ray-Chaudhur respectively proposed a code word that can correct multiple random errors, which is called BCH code, which is outstanding because of its strong ...

AI Technical Summary

Problems solved by technology

However, since the neural network architecture it uses is that the syndrome is used as non-binary information and level information in series and merged directly as the input of the neural network, this will be affected by the RNN structure to a certain extent and affect the decoding effect.

Images