Research On Hybrid Precoding In Millimeter Wave Massive MIMO Systems

With the advent of the 5G era,wireless data services have shown explosive growth,and the data transmission forms have tended to be diversified.Wireless communication has become one of the indispensable technologies in various fields of people's life.However,the scarcity of spectrum resources and the lack of system capacity are the primary problems in the future communication.In order to solve this problem,more and more scholars pay attention to the combination of millimeter wave communication and massive MIMO technology.On the one hand,millimeter waves have abundant frequency band resources,which can provide wider bandwidth and support high-speed data streaming.On the other hand,massive MIMO technology can provide higher antenna gain,thereby compensating for the path loss caused by the shorter wavelength of millimeter waves during transmission.At the same time,the shorter wavelength of the millimeter wave also makes the physical size of the antenna array not too large,which reduces the difficulty of integration.This paper mainly focuses on the design problems of hybrid precoding for the millimeter wave massive MIMO systems,and comprehensively analyzes the performance,the implementation complexity,and the fault tolerance of hybrid precoding schemes.The main contents of this paper are as follows:The millimeter wave communication and massive MIMO technology are briefly described in the introduction.The characteristics of millimeter wave massive MIMO system and the current research status of hybrid precoding technology are described in detail,and the advantages and disadvantages of the existing hybrid precoding schemes are briefly analyzed.The second chapter introduces the millimeter wave massive MIMO system model,the channel model and the precoding technology.Firstly,the common precoding schemes in the traditional MIMO systems are introduced,including the ZF precoding and the MMSE precoding in the linear precoding schemes and the dirty-paper coding scheme in the nonlinear precoding scheme,and making a simulation analysis of their bit error rate.Secondly,the traditional all-digital precoding scheme and the pure analog precoding scheme of the millimeter wave MIMO systems are introduced.Enumerating their characteristics and shortcomings,respectively,and then deriving a hybrid precoding scheme suitable for the millimeter wave massive MIMO system.Finally,making a brief analysis of the two architectures of the millimeter-wave massive MIMO system hybrid precoding,and comparing their advantages and disadvantages by simulation.The third chapter studies the design of full-connected hybrid precoding for millimeter-wave massive MIMO systems.With the goal of maximizing the achievable sum rate of the system,a hybrid precoding scheme based on the PLS-SIC is proposed,and the simulation analysis proves the superiority of this algorithm.The analysis shows that this algorithm can achieve the performance close to the all-digital precoding scheme,is better than the orthogonal matching pursuit algorithm,and is easy to implement.The fourth Chapter studies the joint design of the transmitter hybrid precoder and the receiver combiner for millimeter-wave massive MIMO systems,and proposes a hybrid precoding scheme based on equivalent channels.The scheme is aimed at the architecture of a fully connected receiver and transmitter system,with the goal of maximizing the system spectrum efficiency,processing the analog domain and digital domain of the transceiver,respectively.The analog domain is solved by considering the maximization of channel gain,and the digital domain is solved by constructing a block matrix based on the analog domain to covert the optimization problem into an approximation problem.At the same time,the implementation complexity and the performance of the system when there are large errors in channel estimation are considered.The simulation results show that the performance of the proposed scheme is close to that of the optimal all-digital precoding scheme,the performance and the implementation complexity are better than the orthogonal matching pursuit algorithm and the hybrid precoding algorithm based on channel SVD decomposition.At the same time,when the channel estimation error is 50%,the performance of the proposed scheme can reach more than 70% of the perfect estimation,and it has strong fault tolerance.The fifth chapter is the summary and prospect,analyzes the deficiencies in the article and the direction of follow-up work,and puts forward some open questions about follow-up research.