Block-wise constructing method for quasi-cyclic LDPC code

Abstract

The invention discloses a block-by-block construction method for a quasi-cyclic LDPC code. The invention constructs each block matrix in a quasi-cyclic LDPC code check matrix block by block to cause a ring which has the length of 4 and 6 to not exist between the current constructed block matrix and the prior block matrix by adopting a certain constraint condition, and a (3, L) quasi-cyclic low density parity check code with the minimum ring length being of 8 is finally constructed. Shown by the theoretical analysis that compared with the random construction method of the (3, L) quasi-cyclic LDPC code with the minimum ring length being of 8, the invention greatly lowers the complexity of the construction, and is degraded to a polynomial order from an exponential order. In addition, a simulation result shows that the performance of the quasi-cyclic LDPC code constructed by the invention is superior to the LDPC code based on a random check matrix when short and medium lengths are grouped.

Description

Technical field [0001] The invention belongs to the LDPC code construction in the field of wireless communication technology, and particularly relates to a method for constructing a quasi-cyclic LDPC code check matrix block by block. Background technique [0002] Low-density parity-check (LDPC) codes were first proposed by Gallager in the 1960s. In 1996, Mackay rediscovered that LDPC codes have performance close to the Shannon limit under iterative decoding, which led to extensive research on LDPC codes. [0003] LDPC codes can be divided into LDPC codes based on random check matrix and structured check matrix according to different methods of constructing check matrix. In terms of random check matrix, Gallager and Mackay respectively proposed their own construction schemes. In addition, Chung et al. used a computer to search for the optimal degree distribution based on the density evolution theory established by Richardson, and obtained a randomly constructed LDPC code of 0.0045...

AI Technical Summary

Problems solved by technology

The second type is the LDPC code based on the cyclic permutation matrix. This method simplifies the construction of the regular (J, L) LDPC code check matrix with the row weight L and the column weight J as the selection of JL cyclic permutation matrices. There is a loop in the LDPC code of the permutation matrix. Fossorier gave a random construction method of the QC-LDPC code check matrix based on the minimum ring length of the cyclic permutation matrix. This algorithm requires a full search, so the complexity is extremely high

The third category is LDPC codes based on combination design, which mainly includes LDPC code design based on balanced incomplete block design. The LDPC code of the above-mentioned structured check matrix can realize fast coding, but due to the loss of certain randomness, the group When the length is long, the performance is inferior to the LDPC code based on the random check matrix

[0004] The encoding complexity of LDPC code based on random check matrix is ​​very high, which is not conducive to the realization of hardware

The LDPC code based on the structured check matrix can achieve fast encoding, but due to the loss of certain randomness, when the packet length is long, the performance is inferior to the LDPC code based on the random check matrix.

Images