Abstract:

In order to better realize the reconstruction of sparse check matrix under the condition of high bit error, the process of finding check vectors is converted into the process of solving linear equations, making full use of the sparseness of sparse check matrix. The sparse check matrix is firstly reconstructed by the solution of the sub-linear equation system containing all check bits, and the matrix is used to soft-decode the codeword to correct the erroneous codeword, and continue to reconstruct the sparse check matrix for the corrected codeword. The process is repeated continuously, and the sparse check matrix with the highest reconstruction rate in all the iterative processes is selected as the final result. The simulation results show that the proposed sparse check matrix algorithm can effectively reconstruct the sparse check matrix of bidiagonal structure and non-bidiagonal structure under the condition of high bit error, and has strong robustness to bit errors.

Key words: low-density parity-check (LDPC), linear equations, soft decision decoding