Research And Design Of Spatially Coupled LDPC Code

Spatially Coupled Low-Density Parity-Check(SC-LDPC)codes are able to approach the Shannon limit due to their "threshold saturation" feature,enabling high-capacity communication.Compared to LDPC codes,SC-LDPC codes can achieve the performance of irregular LDPC codes in the waterfall region,but have the low bit error ratio(BER)flat layer characteristics of regular LDPC codes.Burst errors during transmission can lead to deterioration in decoding performance.In this thesis,two different structures are proposed to reduce the possibility of burst errors appearing in the same constraint matrix.The first crossover structure splits the last submatrix of the diagonal band and reorganizes it in the sub-diagonal band direction.The second parallel multidiagonal band structure,which segments the sub-matrices in a single diagonal band of equal length,forms a multi-diagonal band structure.Simulation results show that the SC-LDPC codes of both structures have a certain improvement in decoding performance compared with the conventional diagonal band structure.In order to adapt to real-time channel changes,this thesis proposes two methods for constructing rate compatible SC-LDPC codes in order to achieve code word retransmission at different code rates.The first method is based on the progressive edge growth(PEG)algorithm to construct the parent matrix,the matrix expansion to construct the base matrix with different code rates,the coupling algorithm and the cyclic shift assignment algorithm to form the coupling matrix,and finally the quasi-cyclic(QC)expansion to generate SC-LDPC codes.The rate compatible SC-LDPC codes constructed based on the PEG algorithm have improved decoding performance compared to PEG LDPC codes.The second method: With the help of5 G LDPC codes rate compatible base matrix,the coupling algorithm and cyclic shift assignment algorithm are designed for the special structure of the base matrix,and finally the parity-check matrix is generated.Due to the randomness of the coupling process leading to the emergence of short loops,a short loop elimination algorithm is proposed to reduce its impact on the decoding performance.The simulation results show that the decoding performance of the rate compatible SC-LDPC codes based on 5G encoding structure is better than that of the rate compatible SC-LDPC codes constructed based on the PEG algorithm at all code rates.