Decoding Algorithm Research Of Nonbinary LDPC Code

Low density parity check (LDPC) codes have shown to have good errorcorrecting performance that approaches the Shannon limit, researches on it havedeveloped rapidly in recent years. Binary LDPC codes have been applied in manyways. Nonbinary LDPC codes have better error correcting performance than binaryLDPC codes, but the high decoding complexity has limited its application.Several decoding algorithms are studied in this paper, first studied the decodingalgorithm based on belief propagation (BP), then had research on BP algorithm inlogarithmic domain, the paper also analyzed the extended min sum (EMS) algorithm.Reseach was mainly focused on the Min-max decoding algorithm of nonbinaryLDPC codes. Check node processing as part of Min-max decoding algorithm hashighest complexity, method to lower the complexity was studied in the paper. Layereddecoding scheme was deeply researchd due to it can improve convergence speed. Anew decoding algorithm was proposed in the paper based on Min-max and layereddecoding scheme.Construction method of nonbinary LDPC codes was researched, and twoquasi-cyclic nonbinary LDPC codes were constructed based on the methed.Simulation was carried on the two quasi-cyclic nonbinary LDPC codes using thealgorithm proposed in this paper under Gaussian channel. Simulation results provedthat using the algorithm proposed in this paper not only achieves the low complexityand the low memory requirement, but also speeds up the decoding by two times.