Research On Iterative Signal Detection Algorithm In Massive MIMO

Massive multiple-input multiple-output(massive MIMO)technology is a new generation of mobile communication technology,which can significantly improve the spectrum efficiency and system capacity.The increasing number of users leads to the essential difference between large-scale MIMO and traditional MIMO technology,and a large number of antenna arrays form a complex time-varying channel.The complexity of traditional nonlinear MIMO detection algorithm increases exponentially with the number of antennas,and the performance is poor when used in the 5th Generation Mobile Communication System(5G).In the linear detection algorithm,series expansion method,linear iterative algorithm and gradient approximation algorithm have become the main research objects,but there are still problems of high complexity and low detection performance in the current iterative algorithm.In this paper,several kinds of iterative detection algorithms are mainly studied,and specific improvement suggestions are given.For the problem of slow iterative convergence of Kaczmarz algorithm,the improved R-Kaczmarz algorithm is used to speed up the convergence speed,and the characteristics of R-Kaczmarz algorithm that does not need to calculate the Gram matrix are used to simplify the selection of initial value.When the number of iterations is three,this algorithm reduces the computation by 30% compared with SSOR algorithm and 14% compared with R-Kaczmarz algorithm.Moreover,in order to avoid the slow convergence caused by the sequential iteration of the algorithm,this paper improves the algorithm to be based on the norm ordering,and then iterates.By combining the appropriate relaxation factor,the performance of the improved algorithm is better than that of the original algorithm,with a signal-to-noise ratio of 1-2d B improved.In view of the problem that the detection performance of W-SSOR algorithm is not substantially improved compared with that of SSOR algorithm,the proposed algorithm broadens the selection range of weight factors and improves the performance of bit error rate.Firstly,the convergence of the algorithm is proved.Secondly,Monte Carlo method is used to expand the convergence range of the algorithm and get the value range of the convergence factor.Finally,the simulation results are as follows: the algorithm,while maintaining the same complexity,has a bit error rate performance beyond that of SSOR algorithm and W-SSOR algorithm.