Realization Method Of Polar Encoding And Decoding Based On Linked List

Polar code,also known as polarization code,is the latest channel coding method proposed by Professor Arikan in 2008.The Polar code is a channel coding method capable of achieving channel capacity with strict mathematical proof,and was selected by 3GGP as the coding standard of the control channel in the 5G-eMBB scenario in 2016.The existing theory and simulations show that Polar code has excellent decoding performance and low computational complexity on different channels.In particular,Polar can achieve better performance than traditional LDPC codes and Turbo code in certain scenarios(such as medium and short code length),so it has become one of the most popular research focuses in channel coding in recent years.From the perspective of engineering implementation,this paper implements and optimizes the encoding,channel polarization,channel selection and decoding algorithms of Polar code based on the customized four-way linked list architecture.The main work and innovations are as follows:1.Propose a linked list implementation architecture corresponding to the code tree structure of the Polar code.Different from the traditional matrix or equation-based Polar code implementation method,the linked list structure proposed in this paper sets four pointers,and the linked list nodes closely correspond to the Polar code tree nodes.Based on such a four-way linked list architecture,it can be conveniently and quickly perform engineering simulation and performance testing on channel polarization,coding algorithm and decoding algorithm for different types of Polar codes;2.Based on the four-way linked list architecture of this paper,the channel polarization,channel selection and coding methods of Polar codes are simulated and implemented.At the same time,using the Trellis distribution feature on the linked list,the systematic coding scheme of Polar code is realized based on reverse successive cancellation(SC)decoding.Experimental simulations show that for Polar with a code length of 1024 and a code rate of 0.5,the performance of the systematic Polar code is improved compared with that of the non-systematic Polar code,which is consistent with the simulation results given by Arikan.3.Based on the linked list architecture of this paper,several classical Polar code decoding algorithms are implemented,including successive cancellation decoding algorithm(SC),successive cancellation list decoding algorithm(SCL)and CRC-Aided successive cancellation list decoding algorithm(CA-SCL),and the simulation and performance test of Polar code under different code length and code rate,which provides a reference for the practical implementation of Polar decoder.4.A punctured Polar coding scheme is proposed,which can flexibly change the code length and code rate of the Polar code to achieve the effect of rate matching.We use the LDPC code under the IEEE802.16e standard for comparison and find that the Polar code decoding performance has a 0.5dB gap away from that of the LDPC code at the same code length and code rate.Compared with the quasi-uniform boring scheme proposed by Niu Kai,the equal-division boring scheme has a performance difference of less than 0.1dB.