Research And Design Of Irregular GLDPC Code

Generalized Low-Density Parity-Check(GLDPC)codes are a class of channel codes with a wide range of rate selection and performance approaching the Shannon limit.Compared to Low-Density Parity-Check(LDPC)codes,GLDPC codes replace Single Parity-Check(SPC)nodes with stronger Generalized Constraint(GC)nodes,and their choice of check node component codes is more flexible,resulting in lower error rates under short to medium code lengths and low rates.For the same code length and rate conditions,irregular codes generally outperform regular codes.Therefore,this paper mainly focuses on the research of irregular GLDPC codes.Firstly,to achieve better decoding performance than traditional regular GLDPC codes,a construction method of irregular GLDPC codes based on Progressive Edge-Growth(PEG)algorithm and mixed component codes is proposed.Different row weight mixes in the base matrix are used to construct irregular GLDPC codes with corresponding code lengths of component codes.Secondly,a GC node position selection algorithm based on the number of edges in the Tanner graph is proposed to address the selection problem of GC node positions.Simulation results show that irregular GLDPC codes have significant advantages over regular GLDPC codes in terms of error rate and decoding complexity,and the proposed GC node selection algorithm has obvious optimization effects on both regular and irregular GLDPC codes.In addition,in order to make up for the rate loss caused by the introduction of GC nodes and achieve rate compatibility of GLDPC codes,two puncturing algorithms applicable to GLDPC codes are proposed based on the characteristics of GLDPC codes,namely the group sorting puncturing algorithm based on the number of recovery tree edges of GLDPC codes and the group sorting puncturing algorithm based on the number of GC nodes in the recovery tree of GLDPC codes.The former first groups the variable nodes of GLDPC codes according to the required number of iterations for recovery,and within the same puncture node group,sorts the puncture nodes based on the number of edges calculated for the corresponding puncture nodes in the recovery tree.The latter prioritizes the puncture nodes within the same puncture node group based on the number of surviving GC nodes of the puncture nodes,while also considering the number of surviving SPC nodes and dead check nodes in the recovery tree.Simulation results show that the two puncturing algorithms proposed in this paper outperform random puncturing techniques in terms of error rate and decoding complexity,with the group sorting puncturing algorithm based on the number of GC nodes in the recovery tree of GLDPC codes performing better and being an excellent puncturing algorithm for GLDPC codes.