| The tendency in future mobile communications is higher data transmission rates and higher spectrum utilization.To achieve this goal,the 5G standard adopted many key technologies such as LDPC channel encoding and massive MIMO.LDPC codes has excellent performance close to Shannon capacity.MIMO technology improves spectrum efficiency and increases system capacity.The iterative feedback mechanism jointing LDPC and MIMO makes full use of the performance advantages of both,and improves data transmission rate and communication reliability.Therefore,this paper mainly researches on iterative detection technique based on LDPC codes in MIMO system.In this paper,we first study and analyze various LDPC encoding and decoding algorithms.5G QC-LDPC encoding and minimum-sum decoding algorithm are selected as the encoding and decoding scheme in iterative detection.Then,this paper introduces the MIMO detection algorithms without channel encoding,and compares the complexity and error performance of various MIMO detection algorithms.After that,this paper studies on iterative detection algorithm based on LDPC codes in MIMO system.According to the iterative detection scheme,Jacobi,GS and SOR iterative detection algorithms based on LDPC codes are simulated and analyzed respectively.In the traditional iterative detection algorithm based on LDPC codes in MIMO system,there is a large error in the calculation of log-likelihood ratio when the number of transmitting antennas and receiving antennas is close.To solve this problem,we propose an new method to compute log-likelihood ratios in this paper.This method uses the approximate inverse of tridiagonal component of filtering matrix,which makes full use of the information of filtering matrix and achieves the tradeoff between performance and complexity.The simulation results show that the proposed algorithm can obtain better detection performance.In addition,aiming at the poor reliability of symbol vector feedback to detector in traditional iterative detection algorithm,two improved algorithms are proposed in this paper: one is the iterative detection algorithm based on the average of symbols,which converts log-likelihood ratio information after decoding into average of symbols,the other is the iterative detection algorithm based on symbol adjustment factor,which converts the log-likelihood ratio information after decoding into a symbol adjustment factors.Both of these improved methods can obtain more reliable symbol vectors and improve the error propagation problem in iterative detection by making full use of the log-likelihood ratio information.The simulation analysis shows that the detection performance of the two improved schemes are improved.Finally,this paper studies the design and hardware implementation of 5G LDPC encoder.Aiming at the disadvantages of traditional LDPC encoder based on barrel shift register,such as high complexity and excessive resource consumption,a new 5G LDPC encoder is proposed on the basis of true dual port RAM.The board-level verification results show that the proposed encoder has low complexity,low hardware resource consumption,and achieves 5G full-rate compatibility. |
