Research On Polar Coding Technology For 5G Communication

As a class of channel codes that can provably achieve the symmetric-capacity of binaryinput discrete memoryless channels(B-DMCs),polar codes have won wide attention due to their recursive encoding structure and excellent decoding performance.Recently,research shows that concatenated polar codes such as cyclic redundancy check-polar(CRC-Polar)codes or parity check-polar(PC-Polar)codes can boost the error correcting performance.As a result,polar codes are recommended as the coding scheme for the control channel in 5G enhanced mobile broadband(e MBB)scenario by 3GPP.Concatenated polar coding scheme also becomes a research hotspot.For polar codes with short/medium block lengths,their performance is not only influenced by polarization effect,but also influenced by the minimum distance of polar codes and the multiplicity of codewords with lower code weight,especially in the high SNR region.Considering the development of 5G,we make research on concatenated polar coding technology,analyze and optimize the short concatenated polar coding schemes in terms of distance spectrum in this thesis.The main work is as follows.The fundamentals of polar codes are introduced.The encoding method and the common construction methods of polar codes are described in detail.And decoding algorithms of polar codes in the thesis are also elaborated.The simulations of polar codes with different construction methods and decoding methods are made.The concatenated polar coding schemes and rate matching schemes for 5G e MBB control channel are introduced briefly.The research on concatenated polar codes mainly focuses on CRC-Polar codes and PC-Polar codes.We apply SCL decoding method and ordered statistic decoding(OSD)method on concatenated polar codes and make comparison on the approximate ML performance of several short concatenated polar code schemes.Simulation results show that,compared with OSD,SCL decoding algorithm cannot fully take the advantage of concatenated polar codes.Then the comparison of the distance spectrum and the multiplicity of codewords with lower Hamming weight of these concatenated polar codes are made.Based on the distance spectrum aforementioned,the performance gain of concatenated polar codes is analyzed and the union bound of block error ratio of soft decision ML decoding is calculated.Furthermore,the concatenated polar code scheme is optimized by the improvement of polar code construction and positions of outer codes.As a result,the number of low weight codewords is decreased and the optimized polar codes have better ML performance in the high SNR region.The outer codes can not only improve the distance spectrum of polar codes in the CRC-Polar code scheme and PC-Polar code scheme,but also play a role in error detection.In order to make use of the error correction ability of outer codes,LDPC-Polar codes are investigated.We construct LDPC-Polar codes with different construction methods and apply hard decision decoding method and soft decision decoding method,respectively.Simulation results show that the hard decision decoding performance is influenced by lots of factors such as the code length,the code rate and the error correcting capability of LDPC codes.When considering short length or high code rate,LDPC codes generally have poor performance under hard-decision decoding algorithm.As a result,the LDPC-Polar codes do not perform better than CRC-Polar codes.When the soft decision BP decoding method is used in both the polar codes and LDPC codes,the concatenated LDPC-Polar codes have better performance than the basic polar codes.The error correcting performance can be further improved with SCAN decoding algorithm for inner polar codes decoding.