Research On Encoding And Decoding Algorithm Of LDPC Codes In URLLC

One of the three major application scenarios of 5G,Ultra-Reliable Low Latency Communications(URLLC),is a key communication technology focused on addressing transmission problems that require high reliability and low latency.And the channel coding technology is an important component in ensuring and improving the transmission performance of the URLLC system.Enhanced Mobile Broadband(e MBB),which has been implemented in the three major application scenarios,uses LDPC code for its data channel coding scheme.Even in the "R16" communication protocol frozen in 2020,the data channel coding scheme for URLLC scenario still uses LDPC code.Therefore,exploring LDPC coding and decoding algorithms with superior performance in URLLC scenarios is of great research significance and value.The main research content of this article is as follows:1.The Circulant Permutation Matrix of LDPC codes for URLLC scenarios is constructed.Firstly,we deeply analyze the structure and parameter characteristics of quasi-cyclic LDPC codes under 5G standard.And based on the algebra-Assisted construction algorithm of QCLDPC codes,we analyze the “Cycle Structure” of the check matrix of LDPC codes with Raptor-Like structure.Then,Cycle Structure constraint derived in this paper are added to the base matrix(BG2)which to be used for URLLC.We completed the construction of the circulant permutation matrix of LDPC codes and reduced the complexity of LDPC codes construction process.Finally,the validity and feasibility of the circulant permutation matrix of LDPC codes are verified by simulation.2.An improved algorithm based on self-correcting minimum sum(SCMS)decoding is proposed to meet the demanding time-delay and reliability requirements of URLLC scenarios.This algorithm improves the SCMS algorithm’s approximate calculation process of checking nodes and the judgment process of variable nodes’ reliability threshold.Firstly,the modified threshold of SCMS algorithm is reset according to the influence of check-parity bit information on the judgment of variable node reliability threshold in iterative decoding.Then,the first and second minimums of check-parity bit information are obtained by the normalized minimum sum algorithm,and the normalized factors corresponding to the two minimums are calculated by the order statistics theory.The simulation and complexity analysis results show that the proposed algorithm has excellent decoding performance gain and iterative convergence performance.3.Exploring and learning the ordered statistic decoding(OSD)algorithm with good decoding performance for short codeword length.For the extremely high transmission reliability requirements of URLLC scenarios,an improved ordered statistic decoding algorithm with Cyclic Redundancy Check(CRC)is proposed.The algorithm first replaces the hyperbolic tangent function of the Belief Propagation(BP)algorithm with a specially constructed exponential function,then the CRC code word generating polynomial is transformed into a sub-matrix and integrated into the generating matrix of OSD algorithm.The improved algorithm has excellent decoding performance gain,but it has higher complexity in the higher-order reprocessing process,so the algorithm still needs to be optimized.