Improved Window Decoding of Spatially Coupled LDPC Codes

In this paper, we analyze the performance of window decoding (WD) scheme for spatially coupled low-density parity-check (SC-LDPC) codes, and an improved WD scheme is proposed. Compared with the belief propagation (BP) decoding algorithm, window decoding has the advantages of low decoding latency, low decoding complexity, and small memory. But, the performance of WD is not as good as that of BP. It is found that the error bit number \(\textbf{err}\) of the target symbol does not decrease monotonously with the iteration progresses. This means that the target symbol likelihoods generated when decoding is terminated may not be the best choice to make a decision. To optimize the performance of WD, an improved WD is proposed. In this scheme, the achievable minimum of \(\textbf{err}\) is monitored, and the related likelihoods is stored to estimate the target symbol at the end of decoding. Simulation results show that the improved WD outperforms the conventional WD.