A low-complexity coding scheme for non-binary LDPC code based on IDRB-MLGD algorithm

Non-binary LDPC codes are flourishing in many areas for their excellent error correction performance in last decade. However, the bottleneck for NB-LDPC codes' implementation is the high decoding computational complexity. To seek for a low-complexity decoding algorithm, iterative double-reliability-based majority-logic decoding (IDRB-MLGD) algorithm has been proposed, which still suffers from relatively high error floor at BER level of 10-6. In this work, we propose a low-complexity coding scheme for non-binary LDPC code that bases on iterative IDRB-MLGD algorithm. The proposed work concatenates a Reed Solomon code with non-binary LDPC code which is decoded by IDRB-MLGD algorithm. By such concatenation, errors left by IDRB-MLGD algorithm are set into RS code blocks, which can be further reduced by these RS code blocks. Moreover, a low-complexity decoding algorithm is used to decode the RS code. The comparison results on decoding computational complexity indicate that concatenation only brings about trivial complexity increase. Simulation results show that this proposed concatenated coding scheme can trade off the performance and decoding computational complexity efficiently.