Research On Precoding Technology Of Millimeter Wave Array Antenna Under Limited Feedback

With the rapid development of mobile communication and the diversification of communication services,mobile terminals represented by smart phones have become an indispensable device in human daily life,and the emerging services represented by multimedia is gradually replacing the traditional business with voice message as the main body.The result is a higher transmission rate and better quality of service to meet the growing demand for high-speed communications.However,limited spectrum resources have always been a key factor limiting high-speed data transmission,although existing mobile communication systems have greatly improved spectral efficiency through technologies such as multiple-input multiple-output,orthogonal frequency division multiplexing,and multi-cell cooperation.With the advent of the 5G(fifth generation mobile communication)era,the remaining low-band spectrum resources cannot meet the proposed peak demand of 10 Gbit/s.Therefore,future 5G systems need to find new available resources in higher frequency bands.Millimeter-wave MIMO systems have received widespread attention due to their wider bandwidth than traditional low-frequency MIMO systems and the large amount of spectrum in the millimeter-wave band that has not been authorized.Precoding technology,as a key technical means to enable large-scale MIMO systems to reach or approach the theoretical capacity of systems,is also the focus of research in academia and industry in recent years.This thesis focuses on the precoding technology of millimeter wave MIMO,and considers the hybrid precoding technology of millimeter-wave MIMO system with limited feedback under the condition of unknown channel state information at the transmitter.The thesis mainly focuses on the millimeter-wave MIMO channel and system.First of all,proposed an analog-digital hybrid precoding design algorithm based on finite feedback.Then,for multi-user scenarios,propose a low-complexity two-order hybrid precoding design strategy based on ZF/THP is proposed.The main research achievement as follows:Firstly,the thesis studies the application of analog-digital hybrid precoding in millimeter wave MIMO system,and proposes a design algorithm of analog-digital hybrid precoding based on finite feedback under the condition that the transceiver is not fully known CSI.According to the characteristics of millimeter-wave MIMO system,this method adopts a analog-digital hybrid precoding/decoding architecture,and designs the analog precoding matrix and the digital precoding matrix with the objective function of maximizing the total mutual information.Since the problem is non-convex,the joint optimization problem of analog and digital precoding is simplified to the independent optimization problem of analog and digital precoding.Therefore,considering the finite feedback condition,a hybrid precoding method based on finite feedback codebook is proposed to construct finite analog precoding codebook and digital precoding codebook which are known both on the transmitting end and the receiving end,and a hybrid precoding method is proposed based on finite feedback codebook.The simulation results show that the performance of our proposed algorithm is better than the traditional finite feedback hybrid precoding algorithm.Secondly,millimeter-wave MIMO hybrid precoding technology is applied in multi-user scenarios.Considers the influence of THP precoding technology on system performance,a two-order multi-user hybrid precoding algorithm based on ZF/THP is proposed by adopting the code book construction method mentioned above.Under the power constraint of the base station,the analog precoding matrix,the analog combining matrix the digital precoding matrix,and the digital equalization matrix are designed for each user with the objective total mutual information maximization as the objective function.Finally,the correctness of the theoretical analysis of the proposed algorithm is verified by simulation.At the same time,the simulation results show that although the proposed algorithm has a certain loss of performance compared with the best digital precoding,but it can reduces the computational complexity of the system.Moreover,the performance of the algorithm is better than that of the traditional two-stage hybrid precoding algorithm.