The Low-complexity CA-SCL Decoding Algorithm Based On Path Pruning

Natural attenuation and noise interference in the channel are two important factors that cause inaccurate information transmission.In modern communication system,channel coding technology is widely used to ensure the reliability of information transmission.Polar Code was proposed by Erdal Arikan based on the principle of Channel Polarization.It is the only coding method that can "achieve" Shannon Capacity after strict demonstration among the existing channel coding methods,and has been adopted by 5G standard,and shows broad application prospects in other fields.In this thesis,the construction scheme and decoding principle of Polar code are introduced,and the performance and complexity of existing Polar code decoding algorithms are compared and analyzed.Aiming at the high complexity of decoding algorithm,this thesis makes an improvement based on CA-SCL decoding algorithm.Firstly,the head skipping strategy is used to optimize the decoding process of SCL,and the calculation complexity of a single node is reduced by simplifying the calculation process of path metrics for specific nodes.Second,by judging whether the path is split or not and analyzing the splitting characteristics of the correct path,the number of times of path splitting and re reducing is reduced.Finally,the path pruning optimization scheme is given by combining the head skipping strategy and the splitting reduction strategy,and the Prune-CA-SCL decoding algorithm is proposed.Compared with the mainstream Polar code decoding algorithm,the experimental results show that the Prune-CA-SCL decoding algorithm has the advantage of low complexity,and it is also in the forefront of BER performance,and has good overall performance.When the full polarization ratio is 50%,the decoding performance of Prune-CA-SCL algorithm is 0.1d B lower than that of CA-SCL algorithm,and the complexity is reduced by 60%.