Research On Rate-compatible Polar Coding Scheme

Polar codes,proposed by Arikan,is a new type of channel code which can provably achieve the Shannon capacity limit of the binary-input discrete memoryless symmetric channels via a low-complexity encoding and decoding algorithm.Polar codes have arised a great concern in the academia and been selected for use in the upcoming fifth-generation(5G)communication systems,and they are choosed as the coding scheme under the contorl channel in the Enhanced Mobile Broadband(e MBB)scenario.However,there need accurate channel state information(CSI)to achive efficient coding,and the performance of polar codes is limited when they are ecoded without CSI.Based on the research of polar codes.In this paper,we explored the method to reliase the encoding and transmission of polar codes under the channels without CSI.Moreover,we desigend a rate compatible transmission scheme to realize multiple consecutive polar code blocks by combining the puncturing algorithm.In order to improve the transmission efficiency of each block,we researched a SNR estimation algorithm for polar codes based on the frozen bits error probability under AWGN,and extended to the fading channel to explore the SNR estimation algorithm of the polar code in the fading channel.The main research work is as follows:Firstly,the basic principle of channel polarization is briefly analyzed,and the construction methods of the polar code.At the same time,we researched the classical decoding algorithms of the polar codes,including Successive Cancellation(SC)decoding algorithm and Successive Cancelation(SCL)decoding algorithm.In addition,we investigated an improved random puncturing(IRP)pattern to breack the limitation of blcok length in the power of 2.The proposed IRP algorithm only select puncturing patterns from the frozen bits and keep the information bits unchanged during puncturing,which can achive better performance of punctured polar codes than the exsiting schemes and the construction of polar codes which there are nested information set is realized.Based on the research of exsiting rate compatible polar coding scheme,we proposed a Parallel Concatenated Punctured Polar(PCPP)codes and designed three different transmission schemes on the PCPP codes.The sender transmits polar codes from high rate and low rate until the receiver can decode succesfully.A drawback strategy is used to decode,the receiver decoding from the last polar code,the nested properties of the words are iteratively decoded to decode all polar codes that were previously sent but not successfully decoded until all the information bits are restored to achieve the entire packet transmission.Then,based on the successful transmission of a packet,we designed three different transmission schemes based on PCPP code,by analyzing the overhead of the previous successful decoded PCPP coding block to choose the optimal initial code rate for next block and realized the rateless transmission of multiple blocks of polar codes under stable channel.According to the decoding results of a successful transmission to obtain the current channel capacity,and estimated values of the channel to determine the next packet of the best early code rate.The simulation results show that the three schemes are better than exsting scheme with the same initial rate,and the average number of transmissions is not dependent on the initial channel capacity.Finally,we explored the existing SNR estimation algorithm based on polar code to improve the efficiency of the transmission scheme,and different estimation algorithms are discussed respectively under the infinite code length and finite code length.Then,the algorithm is extended from the AWGN channel to the Rayleigh fading channel and the Rice fading channel.The initial exploration of SNR estimation based on polar code in fading channel is completed.