Reconstruct LDPC codes in a noisy environment

Focus on the problem of reconstructing LDPC codes in a noisy environment, an algorithm based on Column Elimination Operation, Parity-Check Vector Judgment Criterion and Gradual Row Transformation is proposed, which is used to blind recognize LDPC codes. This algorithm first gets enough dual vectors of the receive codes, then selects real parity-check vectors of the dual-space of LDPC codes, and then, detectives and deletes the error code words. The above steps are iteratively carried out until the original problem is reduced to an easier problem, i.e., blind recognition LDPC codes in an error-free context, and finally obtain the sparse parity-check matrix by using Gradual Row Transformation. The simulation and experiment results show that, this algorithm fits most of LDPC standards, including 802.16e, 802.11n, DVB-S2, GJB7296, GB20600, etc., and it is an effective solution for LDPC codes blind recognition in a noisy environment.