Constitution method of binary low density parity check code

Abstract

The invention provides a binary LDPC code constructing method which is based on the PEG algorithm and is provided with a quasi-loop structure. A verification matrix H is symbolized by two small matrixes, one is of a binary basis matrix M the dimensions of which is c*t, and the other is a shaft time matrix P the dimensions of which is c*t. The dimension parameters c and t of the matrix M and the matrix P can be gotten through the code length and the code rate, the basis matrix M can be constructed by the PEG algorithm according to the parameters c and t and the column weight distribution of theverification matrix H, and then the matrix P can be constructed by the right formula. The simulation results prove that the LDPC code constructed by the invention has almost uniform performance withthe random code constructed by the PEG algorithm; and due to the quasi-loop characteristic, decoding with the method provided by the invention is simpler and can be implemented by a feedback shift register. In addition, the code rate is easier to regulate.

Description

technical field [0001] The invention belongs to the technical field of construction of quasi-cyclic low-density parity-check codes, in particular to a construction method of binary low-density parity-check codes. technical background [0002] Low-density parity-check codes, namely LDPC codes, were first proposed by Gallager in 1962. Later, Mackay and Neal et al. re-studied it, and found that the code was decoded by the sum-product decoding algorithm under the AWGN channel. Extreme performance. Many good codes are constructed by random construction methods, but when the code length is large, a large amount of storage space is required to store the check matrix, and since the check matrix has no specific structure, it is difficult to encode effectively. [0003] LDPC code is a kind of linear block code based on sparse check matrix. For a (n, k) binary LDPC code, it can be represented by a very sparse binary check matrix H. If the dimension of the H matrix is ​​m×n and the ra...

AI Technical Summary

Problems solved by technology

When there is a ring, the information sent by a node will be sent back to itself after a ring length of transmission, resulting in the superposition of its own information, destroying the independence of the hypothesis, thus affecting the accuracy of decoding

However, for LDPC codes of finite length, the existence of loops is unavoidable

Images