The Research On Performance Improvement Of Polar Codes With Short To Moderate Length

With the development of mobile Internet and intelligent terminal,the existing 4G mobile communication cannot meet the high demand for communication rate and transmission delay for the mobile Internet applications such as mobile medical,smart city and so on.Therefore,the 5G communication with ultra high speed,low delay and high reliability has gradually become a research hotspot.From the perspective of the physical layer,the high performance channel coding technology is the foundation of reliable transmission in 5G communication system.Polar codes,proposed by E.Arikan in 2009,are known for their ability to achieve the capacity of symmetric binary-input discrete memoryless channels(B-DMC)with low encoding and decoding complexity,which can be well adapted to the 5G communication system with high transmission rate and high reliability characteristics.For Polar codes whose length tends to infinity,its theoretical performance can reach the shannon limit,while for Polar codes with short/finite-length,it have relatively not so good bit error rate(BER)performance as the state-of-the-art channel codes(e.g.Turbo codes,LDPC).Therefore,in order to make the Polar codes applied better in 5G communication,the performance improvement of Polar codes with short to moderate length is studied.The research of this paper includes two aspects: the high performance decoding algorithm and the improvement of Polar codes construction.On the one hand,a good decoding algorithm is the basis for obtaining high performance of Polar codes.In order to improve the decoding performance of Polar codes,the high performance decoding algorithm of Polar codes is studied.The paper introduces the basic decoding algorithm of Polar codes,such as Sccessive Cancellation(SC)decoding algorithm and Belief Propagation(BP)decoding algorithm firstly.However,SC decoding is susceptible to error propagation,and BP decoding has high decoding complexity.Therefore,based on this study,soft cancellation(SCAN)decoding,successive cancellation list(SCL)decoding and cyclic redundancy check-successive cancellation list(CRC-SCL)decoding algorithm of Polar codes are studied,and its decoding performance are also simulated and analyzed.Compared with the existing decoding algorithms,Polar codes have the best performance when adopting CRC-SCL decoding algorithm,and finally the CRC-SCL decoding algorithm is selected as as the basis of subsequent research.On the other hand,the performance improvement of short Polar codes can be obtained by improving the structure of codeword,while concatenating with other codes is the most direct way to change the codeword structure.However,for the existing concatenation schemes,the amount of improvement is largely constrained by the performance of the outer codes with short codeword length.Essentially,one needs to use short outer codes that has performance as high as possible for concatenation to improve the inner short Polar codes performance as much as possible.Motivated by this,in this paper,we propose to use the newly invented Spinal codes,which is characterized by its great error correcting capability over both additive white Gaussian noise(AWGN)and binary symmetric channel(BSC)with short message length,as the outer codes.Due to the channel polarization phenomenon of Polar codes,different bit channels would have different reliability,thus the unequal rate allocation scheme for Spinal-Polar codes is proposed.In addition,in order to reduce the overall decoding complexity,the partial concatenated scheme for Spinal-Polar codes is proposed.Finally,in order to further improve the performance of short Polar codes,we propose a joint iterative decoding algorithm for Spinal-Polar codes,and the decoding complexity is analyzed theoretically.Extensive simulations are carried out,and results show that the proposed concatenation scheme can significantly improve the BER performance of short Polar codes.