Research On Detection Method Based On Neural Networks And Quantum Linear Solvers In Massive MIMO Systems

With the growth of 5G and wireless mobile communication,Massive Multiple Input Multiple Output(Massive MIMO)technology has become a hot research topic in the communication field.However,with the increasing size of the antenna array,it makes the design of the receiver more difficult,and the detection algorithm with too much complexity is increasingly difficult to be implemented in practical engineering.Therefore,how to design a detection algorithm with excellent detection performance and low operational complexity in Massive MIMO systems is the key to the development of this technology.In this problem,the main research work of this thesis includes.(1)The thesis firstly studies and analyzes the system model and related signal detection algorithms based on the Massive MIMO uplink system,addressing the maximum likelihood estimation detection for optimal performance,the suboptimal linear detection,nonlinear detection algorithms,and the neural network model-driven detection algorithms,respectively.(2)The thesis unfolds a high-parallelism detection algorithm(HP)into a neural network,constructs the corresponding network framework based on a single-layer neural network for each iteration of this algorithm,and combines it with trainable weighted parameters and nonlinear neural units to propose a network structure based on the HP-Net(High-Parallelism Net,HP-Net)based detection scheme.The network model with optimal weighting parameters is obtained by training HP-Net,and then the detection performance is improved.The experimental results show that the thesis method has lower complexity and lower BER than the traditional Minimum Mean Square Error(MMSE)algorithm,and better BER performance than the HP parallelism detection algorithm.(3)The thesis proposes a signal detection algorithm based on quantum linear solver to address the problem of exponential growth of operational complexity of traditional signal detection algorithms with the number of transmitting antennas,using the natural parallel nature of quantum computing.By taking advantage of quantum computing,the HHL(Harrow-Hassidim-Lloyd,HHL)algorithm is applied to the detection of Massive MIMO systems,the received signal is mapped with quantum amplitude encoding,and the detection results are solved and extracted by quantum linear solver and quantum laminar technique,and the Massive MIMO detection method based on quantum linear solver is proposed.detection method based on quantum linear solver.The theoretical and simulation results show that,compared with the traditional MMSE algorithm,the method effectively reduces the complexity of the algorithm and achieves exponential acceleration of the computation while ensuring the performance.