Research On Improved Decoding Algorithms Of Polar Codes Over The AWGN Channel

In 2008,E.Arikan proposed the concept of channel polarization and based on this he provided a new kind of codes,namely polar codes,which can reach the Shannon limit under symmetric binary-input discrete memoryless channels(B-DMC).The Polar code has good performance and can be encoded and decoded with low complexity.Therefore it is a major breakthrough in coding theory and becomes one of the research hotspots.By improving decoding algorithms,the polar code could achieve better performance over all kinds of channels.This paper focused on polar codes and studied their decoding algorithms over the B-AWGN channel.The contributions of this paper are shown as follows:First,this paper studied the error probability of sub-channels selected by the improved gaussian approximation.Since for polar codes with finite length the selected message bits in sub-channels have a certain error probability,we exploited the resources of sub-channels thoroughly,analyzed the statistic of the error probability of message bits in sub-channels over the B-AWGN channel,and obtained the distribution of the error probability.This laid the foundation for proposing the improved decoding algorithm.Secondly,we proposed the Single Parity Check Code Based Subsection Check Aided Successive Cancellation List(SPCC-SC SCL)algorithm according to the distribution of the error probability derived above.SPCC-SC SCL inserts the single parity check code into subchannels and performs subsection check during decoding to select the good path.Simulation results show that compared with CRC-aided SCL using CRC-24,without degrading performance,SPCC-SC SCL achieves lower complexity and provides better error performance in low signal to noise ratio(SNR)region.