Decoding method and device, decoder

Abstract

Embodiments of the present invention provide a decoding method and device, and a decoder. The method includes: receiving N LLRs corresponding to the signal to be decoded, dividing the K decoding bits into reliable bits and unreliable bits according to a priori LLR and / or a posteriori LLR, and generating M according to the N LLRs and preset rules There are 1 decoding paths, the next-level path for reliable bits is 1, and the next-level paths for unreliable bits are 2. According to the PM value of the M decoding paths, the target decoding paths of each level are screened out, and each target decoding path is obtained. When the length of the first decoding path reaches N, perform CRC on the K decoding bits in the first decoding path, if the CRC passes, the decoding is successful; otherwise Continue to perform CRC on the K decoding bits in the next decoding path whose length reaches N. If the result of performing CRC on all K decoding bits in the decoding path whose length reaches N is not passed, then decoding fail.

Description

technical field [0001] The present invention relates to the technical field of coding and decoding, and in particular to a decoding method and device, and a decoder. Background technique [0002] Polar (polarization) codes are a coding method that is theoretically proven to reach Shannon's limit. Polar code is a kind of linear block code, and its encoding process is in is an N-length binary row vector, B N is a N×N bit inversion (Bit Reversal) matrix, for log 2 N matrix F 2 The Kronecker product of , Some of the bits are used for information transmission, called information bits, and the remaining bits are set as fixed bits. Polar codes have excellent decoding performance in a wide range of work (code length, code rate, signal-to-noise ratio). [0003] The decoding process of the existing bit-by-bit cancellation decoding algorithm (Successive Cancellation, referred to as: SC) is: after receiving the signal, calculate the log likelihood ratio (Log Likelihood Rat...

AI Technical Summary

Problems solved by technology

[0004] In practical applications, the number of decoding bits is very large. Using the SCL decoding algorithm, although the decoding performance is better, it avoids the defect that a decoding bit error in the SC algorithm will spread to subsequent decoding bits. , but to calculate the PM values ​​of all paths, for each decoding bit, all paths must be sorted according to the PM values, and the computational complexity and sorting complexity are very high

Images