Qc-ldpc Codes In Optical Fiber Communication Applications

Low-Density Parity-Check (LDPC) codes are a class of capacity approaching error-correcting codes and a good liner block codes, which perform at a rate extremely close to Shannon capacity. Its error-correcting ability has tight connection with the construction way in encoder. QC-LDPC codes are characterized by the parity-check matrix which consists of small square blocks which are the zero matrix or circulate permutation matrices. It is an important branch of LDPC codes. Encoding of a quasi-cyclic (QC) LDPC code can be achieved with simple shift-registers, being prone to hardware realization.In recent years, LDPC codes are using in the fields of mobile communication, satellite communications and magnetic storage system because of its excellent performance. It can improve the optical communication system transmission performance if the system using LDPC codes as FEC (Forward Error Correction) Code. However, we should consider the fiber communication system's character, eg. high rate and high efficiency.Quasi-cyclic low-density parity-check (QC-LDPC) code can achieve higher coding gain for fiber optic communication systems, in this paper, the main contents and results are as follows.The major factors such as fiber channel model, decoding algorithm and iterations which impact QC-LDPC codes' performance are simulated. But the computation of log-likelihood ratio (LLR) in the decoding process is very complicated for fiber channels. To reduce the complexity of the decoder hardware design, this paper introduces a novel simplifier method for the LLR computation in fiber channels. Through simulation results with min-sum decoding algorithm, it is found that the simplifier method has a performance loss of only 0.05dB when compared with the original LLR computation method.Quantization decoding of QC-LDPC codes is studied. We simulated the decoding performance which is affected by quantitative precision and the scope of quantization. The simplifier LLR with 3 bits quantization is simulated and the result shows it just lost 0.002dB performance compared with the original LLR when the decoding algorithm is MS.In the end, we proposed a novel decoding method for improving the decoding performance, this new method can achieve 0.4dB decoding gain compared with the original method when using MS algorithm. Using simplify LLR in this method is essential to simulated.