Viterbi decoding based multi-path parallel loop block back tracing method

Abstract

A multi-channel parallel loop block backtracking method based on Viterbi decoding is based on the principle of Viterbi decoding. The process includes: proposing a decoding method that only needs to use convolutional code coding backtracking, and the backtracking method decodes the survivor path information of Viterbi decoding Depending on the situation, single-port RAM or dual-port RAM can be used for storage, and the number of RAM units is always the product of twice the backtracking depth T and the total number of convolutional code states. RAM needs to be divided into multiple blocks, and the number of blocks is determined by the backtracking depth and the size of the backtracking block. Each block of RAM corresponds to a backtracking unit and a set of backtracking gating network. Only a part of RAM needs to be read and written during backtracking. At the same time, the number of effective parallel backtracking paths is the ratio of backtracking depth to backtracking block size. During the backtracking process, the starting point is the state value corresponding to the minimum value of the path metric as the current backtracking state. Then, the current backtracking state is combined with the selected current surviving path to backtrack to the next decoding state value, and multiple blocks of RAM need to be continuously The current status value of the access and backtracking is also continuously switched between the two RAMs. Finally, the backtracking results are arranged in reverse order, and the decoding results are output in the normal order after being registered.

Description

technical field [0001] A Viterbi decoding scheme for convolutional error correction coding - a multi-channel parallel loop block backtracking method, relates to signal processing in communication chips, belongs to the field of wireless communication, and especially relates to area and power consumption optimization for chip design. Background technique [0002] As a forward error correction coding technology, the product code is widely used in wireless communication, mobile communication, deep space communication, and disk storage error correction. However, the decoding of convolutional codes is difficult, and two known methods are algebraic decoding and probability decoding. Algebraic decoding is based on the encoding structure of the convolutional code itself, and the statistical characteristics of the channel are not considered during decoding. Probabilistic decoding should consider the statistical characteristics of the channel when calculating. Typical algorithms such...

AI Technical Summary

Problems solved by technology

The register exchange algorithm occupies a huge area and consumes a lot of power, so it is generally not used except for occasions where the code rate is particularly high.

Images