LDPC-based Turbo structure code hybrid concatenated encoding and decoding method

Abstract

The invention discloses a novel channel coding, which is an LDPC-based Turbo structure code hybrid concatenated encoding and decoding method. The Turbo structure code firstly utilizes LDPC as a component code, and after being encoded by an LDPC outer code encoder, the Turbo structure code is divided into two mutually parallel paths by an interleaver, wherein one path is output after being processed by an LDPC inner code encoder, and the other path is output after being in turn processed by the interleaver and the LDPC inner code encoder. According to the LDPC-based Turbo structure code hybrid concatenated encoding and decoding method, a hybrid concatenated encoding manner and an interweaving technology are utilized, and the LDPC is utilized as the component encode, so that a code word has an stochastic approximation character; great decoding performances can be obtained in low SNR, and the traditional parallel concatenated code error floor can be effectively eliminated; the input and output of a decoding module of each component code are soft information, so that the algorithm provided by the invention can obtain higher encoding gain.

Description

(1) Technical field [0001] The invention belongs to the field of channel coding, and in particular relates to a novel coding and decoding method of a hybrid concatenated turbo structure code using LDPC as a component code. (2) Background technology [0002] Shannon's theorem is the basic theory of channel coding. The theory points out that, on the premise that the channel transmission rate R is less than or equal to the channel capacity C, the random edge decoding method is adopted, and the maximum likelihood decoding algorithm is adopted at the receiving end, the encoding length is long enough and the decoding is optimal. The maximum likelihood decoding method can make the bit error rate close to zero. [0003] In order to find a good code that satisfies Shannon's condition, people have made a lot of efforts to develop concatenated codes, soft-decision decoding and various modulation methods, but the best coding gains of the above methods are also 2~3dB below the theoretic...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to propose a kind of error floor (errorfloor) problem that can occur when the BER of traditional parallel turbo codes drops to a certain extent with the increase of SNR, or even no longer decreases, and traditional parallel turbo codes. Turbo's component codes mostly use recursive system convolutional codes, resulting in high requirements for signal-to-noise ratio and low coding gain, which cannot meet the application problems when channel conditions are poor.

Images