Encoding And Decoding Algorithms Of Fountain Codes

Fountain codes have advantages of robustness and rateless. It has been successfully applied in the FEC(Forward Error Correct), DC(Data Compression), DS(Data Storage) fields and so on. The optimization of encoding and decoding is a research hotspot in Fountain codes.Robust solitary waves in the distribution of degree-type control and pre-coding technology can effectively improve the decoding performance of Fountain codes and solve the rateless transmission problem in the Erasure channel. However, R assemblage, short-loop, DI(Discrepancy Information), the noise problems directly affect the cost of decoding and application environments of the Fountain codes. This paper gives a systematic investigation of CDI(Control the Degree of Information unit), ESL (Eliminate the Short Loop) , DI , BP (Belief Propagation) algorithm.First of all we summarize the degree distribution, encoding and decoding process, decoding efficiency of LT(Luby-Transform) codes and Raptor codes. Because the R assemblage with empty will lead to decoding failure. We propose the SMI(Select the Minimum degree of Information unit) algorithm to reduce the probability of R assemblage with empty. On the other hand the SMI algorithm could increase the short-loop that lead to decoding failure. Proposing an optimization scheme to eliminate the short-loop which the length is 4 or 6. The simulation results show that the new algorithm reduce the failure probability of decoding.Based on the analysis of DI problems of decoding process of Fountian codes, it come to the conclusion that the decoding efficiency can be improved by releasing the DI unit. Proposing an optimization algorithm for DI and getting a new algorithm, which used the check unit to simplify decoding process.Researching the performance of the Fountain codes transmission in AWGN(Additive White Gaussian Noise) channel by BP(Belief propagation) transmission algorithm which use the information iterative decoding algorithm on this paper. Using ESL, CDI and simplify decoding algorithm to improve on the decoding process. Experiments show that the improvement algorithm not only improves the decoding efficiency ,but also reduces the computation complexity.