Research On The Encoding And Decoding Of Non-Binary LDPC Codes In High-Speed Optical Transmission Systems

With the coming era of Big Data, the demand for network bandwidth and data traffic is increasing, while the rapid development of high-speed optical transmission system, the transmission effect of fiber is a serious impact on the system to further enhance the rate and transmission distance. As a key factor in solving all physical damage, the forward error correction technology which can protect system reliability and increase the transmission distance is becoming the hotspot.LDPC codes as the third generation soft decision technology which is extensively studied, has a superior error correction performance, low error floor and can be implemented in parallel. Compared with binary LDPC codes, non-binary LDPC codes have a stronger burst error and random error correcting capability, in addition it is more attuned with the higher-order modulation and more suitable for ultra high-speed long-haul optical transmission systems. However, the study of non-binary LDPC codes is less, the theory is not perfect, and the high complexity of its decoding algorithm seriously affecting its massive popularity. In this paper, I will focus on non-binary LDPC codes, mainly to study the encoding and decoding algorithms. A series of simulation analysis will be conducted based on128Gb/s QPSK system. The specific contents and findings are as follows:(1)The basic principles of LDPC codes is studed firstly, and research is focused on constructors and decoding algorithms of binary and non-binary LDPC codes, detailed analysis of The specific implementation steps are analyzed in detail. All of these are theoretical preparations for MATLAB Simulation.(2)Research on coded modulation technique. The basic principles of block-interleaved coded modulation, polarization-multiplexed coded-OFDM and nonbinary LDPC-coded modulation are then introduced. In order to provide some application ways of LDPC codes used in high-speed optical transmission systems.(3) A detailed description of the Eb/NO calculation method in MATLAB simulation, to make sure that the error correction performance can be evaluated visually.(4) Set up a128Gb/s QPSK simulation system based on NB-LDPC-CM, and make a series of simulations about error correction performance between GF(4) NB-LDPC codes and binary LDPC codes. The results showed that the error correction performance of4-ary LDPC codes is binary LDPC’s50times at Eb/NO is3.3dB; and the NCG of4-ary LDPC code is also increased about0.25dB than that of binary LDPC code at BER is10-4.(5) Analyse the influence of iterations, code length and rate of non-binary LDPC codes for error correction performance by simulations. The results show that the optimal number of iterations for4-ary LDPC codes using MD-FFT-QSPA decoding algorithm is50, and its error correction performance increases with increasing code length, decreases with increasing code rate.(6) A decoding algorithm which involve flag is proposed for non-binary LDPC codes. The simulation results show that, the NCG of adding flags algorithm is about0.25dB higher than that of traditional decoding algorithm at BER is10-6; at the same time, the new algorithm also can effectively reduce the decoding iterations and power consumption.