| The fifth-generation mobile communication technology(5G)promotes the rapid development of the wireless communication industry,but the contradiction between the explosive increase of wireless network users and the lack of spectrum resources in the wireless communication system is increasingly prominent.Millimeter wave frequency band has a large number of wireless resources that have not been effectively utilized.Therefore,the millimeter wave is becoming an important medium for 5G signal transmission.multiple-input multiple-output(MIMO)technology can provide enough beam gain to offset the very serious channel weakening in millimeter wave channels.Therefore,the millimeter wave is selected as the main signal carrier in 5G.Then the millimeter wave is combined with massive MIMO technology to ensure the safety and high efficiency of the wireless communication system.At the same time,as a necessary signal processing technology in massive MIMO systems,hybrid precoding technology can achieve a flexible compromise between energy efficiency and spectral efficiency in wireless communication systems.Given this,this thesis mainly studies the hybrid precoding technology under the partial connection architecture.Two new precoding schemes are proposed in this thesis.At the same time,we analyze the system performance of the proposed scheme,such as spectrum efficiency,energy efficiency,implementation complexity,and robustness.The specific research contents are as follows:(1)Aiming at the problem of poor system performance in existing partially connected hybrid precoding schemes,an alternative optimized hybrid precoding scheme based on initial value acceleration is proposed in this thesis.Firstly,aiming to maximize the trace of the product of the optimal full digital precoding matrix and mixed precoding matrix,the digital precoding matrix was designed by utilizing singular value decomposition(SVD),and then the analog precoding matrix was optimized and updated in the process of alternate optimization.In addition,a heuristic algorithm is proposed to improve the channel gain.By solving the initial value of the iteration as the starting point of the alternating optimization,the algorithm can effectively reduce the probability that the alternating result converges too slowly to fall into the suboptimal solution.The simulation results show that the spectral efficiency of the proposed algorithm is better than that of other partial connection precoding schemes when the complexity is increased only a little.(2)To reduce hardware complexity and power consumption in massive millimeter-wave MIMO systems,previous precoding efforts have focused mostly on fixed partially-connected architectures.However,the diagonal constraints of analog precoders cause a lot of performance loss to the system.To solve the above problems,a dynamic subarray architecture is proposed to realize the tradeoff between spectral efficiency and energy efficiency in the system.And the corresponding hybrid precoding algorithm is designed.Specifically,this thesis decouples the multivariable joint optimization problem into a simple maximum sum rate optimization problem.Furthermore,an alternate minimization algorithm is proposed to calculate the analog precoding matrix and complete the effective partitioning of transmitting antennas on the RF chain.Then the least square method is used to solve the digital precoding matrix.Simulation results show that the spectral efficiency of the proposed scheme gets a good compromise between the fully connected structure and the fixed partially connected structure.The simulation results show that the spectral efficiency of the proposed scheme achieves a good balance between the fully connected structure and the fixed part connected structure.And the scheme has high energy efficiency.In addition,it can guarantee the system’s performance under the condition of poor channel information and has strong robustness. |
