Research On Signal Detection Algorithm In Massive MIMO System

Because of the development of mobile devices and mobile services,there will be huge challenges in terms of channel resources,transmission performance,and channel capacity for communications in the future.Also,Massive multiple-input multiple-output(MIMO)technology can improve channel capacity and spectrum utilization,it has been wildly used today.Although Massive MIMO has so many advantages,there are still a series of problems behind it.For example,there are a large number of antennas are deployed on the base station,as a result,the dimension of the channel matrix is so high that the received signal is more complicated.Therefore,on the basis of previous research,this thesis has carried out further research on the signal detection algorithm in the massive MIMO system.For example,likelihood ascent search(LAS)and reactive tabu search(RTS).The main research of this thesis is as follows:1.For a likelihood ascent search algorithm,its maximum likelihood solution is not necessarily located in the neighborhood.Also not all the ML cost function values of the solution vector are smaller than the current solution vector,that is,the search process may go through for a large number of intermediate vectors,an improved signal detection algorithm based on LAS is designed.The improved algorithm combines the idea of reduced neighborhood,and decides whether to include the vector in the neighborhood through a metric and some selection rules.This algorithm can reduce the complexity of the algorithm effectively by reducing the neighborhood vector.At the same time,it can also improve the operating speed of the system to a certain extent.In the end,the correctness of the proposed algorithm is verified by matlab simulation.2.Because of the poor performance of symmetric successive over-relaxation reactive tabu search(SSORRTS)algorithm in high-order modulation systems,this thesis proposes a SSORRTS-BP detection algorithm combined with the belief propagation(BP)algorithm.In view of the high complexity of the RTS algorithm,some scholars propose to improve its initial solution performance.This algorithm preprocesses the initial solution of the RTS to avoid matrix inversion operations.So it can reduce the complexity of the algorithm effectively.However,this algorithm cannot solve the problem of poor algorithm performance in high-order modulation systems.So the SSORRTS-BP hybrid algorithm is designed to solve this problem.The SSORRTS-BP algorithm uses the output of the SSORRTS algorithm as the input of the SSORRTS-BP hybrid algorithm.At the same time,the BP algorithm is used to correct the least significant bit error of the SSORRTS.At the end,The simulation experiment shows that the SSORRTS-BP algorithm can improve the performance of SSORRTS detection in high-order modulation systems,also reduced the bit error rate of the system.