Research And Simulation Design Of Polar Code Coding And Decoding Algorithm Based On Layerde Bit-flipping

Polar code was first proposed by professor Arikan in 2009.It is the only coding scheme that can strictly prove the channel capacity of the binary input symmetric discrete memoryless channel.Polar code has become a highly concerned channel coding technology due to its low encoding and decoding complexity and has been adopted as a 5G-e MBB control channel coding standard.This paper mainly studies and simulates Polar code decoding algorithm of short and medium code length.The main contents and innovations are as follows:1.Polar code channel polarization methods are studied and analyzed,including channel combining,channel splitting,channel polarization and so on.Channel polarization phenomena are analyzed through specific examples and data.As for Polar code coding and channel selection,this paper estimates the reliability of Polar channel based on monte carlo simulation,density evolution tool and gaussian approximation theory,and analyzes and describes the whole process of Polar coding with concrete examples.2.On the basis of successive cancellation(SC)decoding,the successive cancellation list(SCL)decoding algorithm is studied and analyzed in combination with the concepts of code tree structure and path measurement.Furthermore,the CA-SCL decoding algorithm with cyclic redundancy check(CRC)code is discussed and analyzed.Finally,based on the comprehensive simulation platform built in this paper,the decoding is realized by using the linked list method,and the complexity and performance are analyzed and compared based on the specific simulation data.3.Based on the phenomenon of error propagation,Gauss Approximation(GA)method is adopted to select the information bit that is most prone to error,so as to improve the traditional successive cancellation single-bit-flipping decoding algorithm(SCSF).On this basis,a Layered Cyclic Redundancy Check-based SCL(L-SCLSF)algorithm is proposed.In this algorithm,the idea of bit flip is applied to SCL,and the information bits are layered,and each layer is CRC checked to achieve the effect of flipping multiple bits at once,so as to improve the decoding performance.When the number of flips is 32,the performance of SCSF decoding based on gaussian approximation is slightly better than that of CA-SCL decoding algorithm with 2 search width.At the same time,the performance of L-SCLSF algorithm is better than that of SCL and CA-SCL decoding algorithms with the same code length and search width.The performance gains are obvious under short code length or high SNR.Polar code coding and decoding integrated simulation platform based on linked list was designed and completed based on C++ language.According to the butterfly symmetrical structure of Polar code,the platform realizes the encoding and decoding process of Polar code based on four-way linked list and self-defined pointer structure.It avoids complex matrix calculation and improve the efficiency and stability of software simulation.