Research And Improvement Of Polar Code Coding And Decoding Algorithms

With the development of mobile networks,people have higher requirements for the communication quality.It is important to choose a reliable channel coding scheme.Professor Erdal Arian from Turkey proposed the polar code in 2009,which has attracted more attention about channel coding in academia.Polar code can be theoretically proved to reach the Shannon limit.Polar code can reach the Shannon limit and has low complexity for long code bits as an important coding scheme in 5G wireless systems.However,the decoding performance of short code bits is not ideal due to the insignificant channel polarization.In this thesis,we focus on the decoding algorithm of the polar code and improve the decoding technology to optimize the decoding performance.The main work of this thesis is as follows:(1)The Successive Cancellation decoding algorithm of polar code is analyzed detailedly in Chapter 3.Due to the frozen bit is known to the transmitter and the receiver,the traditional Successive Cancellation algorithm has some redundant computation.By pruning the computation of the frozen bit,this thesis proposed an improved Successive Cancellation decoding algorithm,which effectively reduces the complexity of the traditional Successive Cancellation decoding algorithm.Simulation results show that the proposed strategy can efficiently reduce the complexity by nearly 22% when the code rate is 0.5.(2)Based on the improved Successive Cancellation algorithm in Chapter 3,this thesis proposed Single-Bit-Flipping Successive Cancellation algorithm with Cyclic Redundancy Check in Chapter 4.In other words,the introduced Cyclic Redundancy Check code and polar code constitute a simple concatenated code,which can further improve the error correction capability of the decoding algorithm while minimizing decoding complexity.Simulation results show that the block error rate of Single-Bit-Flipping Successive Cancellation algorithm based on Cyclic Redundancy Check with 16 flips has good performance gain about 0.5d B.(3)This thesis analyzes the shortcomings of the concatenation scheme formed by polar code and LDPC in Chapter 5.The concatenation scheme has no strong error correction ability for continuous errors.We introduce matrix interleaver to convert continuous errors into random error code,the scheme can improve the error correction capability of the concatenation scheme for continuous errors.The decoding scheme uses Belief Propagation algorithm and Successive Cancellation List algorithm,which can effectively avoid the error propagation of Successive Cancellation algorithm.Compared with the unimproved concatenation scheme,the performance of the concatenation scheme with matrix interleaver can be well improved.