Research On Implementation Of LDPC Code In 5G Data Channel

The Low-Density Parity Check(LDPC)code is one of the error correction codes whose performance approaches the Shannon limit.Due to its superior performance,it has become the coding scheme for the data channel of enhanced Mobile Broadband(eMBB)in the 5G mobile communication.Therefore,it's of significance to research on low-complexity coding and decoding,low-delay coding and decoding and high-throughput of LDPC codes.The specific research contents are as follows:1.A recursive coding algorithm is proposed for LDPC codes in data channel.In this thesis,LDPC code is simulated and analyzed in data channel and the decoding performances of belief propagation(BP)algorithm,normalized min-sum(NMS)algorithm and offset min-sum(OMS)algorithm are compared.In order to facilitate engineering implementation,the fixed-point quantization scheme and iteration times are determined through the fixed-point simulation,and verified the algorithm.2.The convergence speed of belief propagation decoding based on serial scheduling strategy is analyzed in the thesis.The group of variable nodes is fixed in the general shuffled belief propagation algorithm(SBP),and the transfer information in sub-iterative decoding cannot be fully utilized.Therefore,a dynamic adaptive group shuffled belief propagation algorithm(AGSBP)algorithm is proposed in this thesis.The main idea of AGSBP is to determine the update order of variable nodes according to the reliability of variable nodes.The simulation results show that AGSBP has better decoding performance and faster decoding convergence speed.3.Compared to the parallel scheduling algorithm,although the dynamic serial scheduling algorithm improves the decoding convergence speed,there are a large number of intermediate operations in the decoding process,which is difficult to achieve in hardware.Therefore,a fixed-scheduling layered belief propagation(LBP)algorithm is designed,which is an approximation of the credibility of variable nodes.The credibility of the variable nodes affects the information update of the check nodes.when updating the check nodes in every layer,variable nodes with low credibility need to be included as many as possible.The fixed scheduling decoding algorithm determines the check nodes update order before decoding,so it does not increase the complexity of decoding.The simulation results show that the serial scheduling algorithm designed in the thesis can effectively improve the decoding performance and speed up the decoding convergence speed.