Research On Optimization Of Polar Codes Decoder

Polar code is a new channel coding method proposed by E.Arikan in 2009,it is the only one that has been certified to achieve the Shannon limit memory channel when code length is long enough in case of binary discrete memoryless channel.The polar code is based on polarization phenomenon,it use the combination of channel encoding and the channel separation structure.Polar code in the case of infinite length to reach the Shannon limit,in the actual communication unlimited length is impossible.Therefore,it is necessary to search for an efficient decoding algorithm under the limitation of the limited code length.In this paper,an optimization scheme is proposed based on the SC decoding algorithm of the polar code,the scheme improves the judgment condition of the information bits in decoding to the LR value comparison result of the next information bit in the current information bit.The main contents of this paper are as follows:(1)Analyze the basic principle of polar code.The channel combination,channel splitting and channel polarization are discussed in detail.The polarization mode in the binary Gauss channel is discussed,and the corresponding parameter calculation method is given.(2)The coding and decoding of polar code are studied in detail.The encoding process of polar code is more intuitive,the most important part of coding is to obtain the generation matrix and channel selection.In the decoding,the algorithm of SC is analyzed in detail,and it is based on the recursive method,which is based on the LR value.The SC decoding algorithm is simulated in the Gauss channel,and the decoding performance is analyzed,and put forward the performance of the SC decoder is not high when the code length is limited.(3)The optimization scheme is proposed based on SC decoding algorithm,and the design of decoding is given.In the decoding process,the optimization algorithm to determine the current information bits need to be based on the next information in the current bits 0 and1 under the conditions of the LR value comparison,this method reduces the error rate of the decoding.Through the comparison of the improved algorithm and the SC decoding algorithm,the complexity of the decoding algorithm is given.The simulation results show that the bit error rate of optimized decoding is lower than that of SC,and the error rate curve is analyzed.