Research On Massive MIMO Technology In Mobile Communication Systems

With the explosive growth of mobile data services and user access in recent years,the advantages of high spectral efficiency and high energy efficiency of massive multiple-input multiple-output(MIMO)technology have been generally recognized by academia and industry.Massive MIMO has been widely recognized as one of the key technologies for 5th generation mobile communication(5G).However,the implementation and deployment of massive MIMO still face certain challenges,such as feedback compression for channel state information(CSI)in massive MIMO systems,beamforming in massive MIMO with imperfect CSI,activation of partial transceivers and coordinated beamforming(CoBF)in heterogeneous networks,etc.In addition,the convergence of massive MIMO and non-orthogonal multiple access(NOMA)technologies is also a problem worth studying.Based on the above problems,this thesis studies the CSI compression feedback technology for massive MIMO systems and the hybrid beamforming for massive MIMO systems with imperfect CSI feedback.Then,relevant research has been taken in remote radio head(RRH)activation and robust CoBF for massive MIMO heterogeneous cloud radio access net-works.Finally,the downlink performance of the massive MIMO NOMA systems under actual channel estimation has been studied and evaluated.Specifically,the contributions are summarized as follows:Firstly,the channel matrix dimension of massive MIMO is very high,direct feedback of the estimated channel information will occupy a large amount of spectrum resources,resulting in a significant decrease in the effective capacity of massive MIMO systems.Fortunately,existing research shows that massive MIMO channels have a strong spatial correlation due to the limited physical distance between antenna elements.Based on the property that high-dimensional strong correlation data falling into a low-dimensional affine hull,an affine set fitting(ASF)-based approach is proposed to reduce the CSI feedback overload for massive MIMO systems.Moreover,the complexity of the proposed algorithm is lower than that of the two-dimensional discrete cosine transform and the Karhunen-Loeve transform based algorithms,and it is especially suitable for the realistic scenario where the computing capability of the user equipment is limited.In addition,the proposed ASF algorithm has the same performance as the classical principal component analysis and has the ability to suppress noise.Since the limited channel compression feedback will cause a certain loss of precision,and the non-ideal estimation of the channel and the feedback delay,it is unrealistic for the transmitter to obtain perfect CSI of each user.To tackle this problem,the hybrid beamforming of minimizing CSI mean square error is studied by substituting the analog domain and the digital domain beamforming matrix variables with a full-digital domain beamforming matrix,then applying semidefinite relaxation(SDR)to transform the original problem into a semidefinite programming(SDP)problem,and proposed an alternate iterative robust hybrid beamforming algorithm to obtain the optimal full-digital domain beamforming matrix.Finally,the two-stage iterative algorithm is used to obtain the analog domain and the digital domain beamforming matrix of the original problem,respectively.Simulations are provided to validate that the proposed algorithm can provide superior performance to that of the nonrobust algorithm at different channel error and signal-to-noise ratio scenarios.Secondly,considering the imperfect CSI as well as non-negligible of the power consumption in fronthaul links at the RRHs and in the RF circuits,we propose a joint RRH activation and outage constrained CoBF algorithm for massive MIMO heterogeneous cloud radio access networks.In order to solve the nonconvex joint RRH activation and robust CoBF problem,we first study the beamforming design problem with a given activated RRH set.Conservative convex approximations for the outage constraints of RRH user equipments are derived by using SDR and an extended Bernstein-type inequality,while a closed-form expression is obtained for the outage constraints of macro base station user equipments.Furthermore,we reformulate the nonconvex problem into an SDP.Then,we propose a low-complexity algorithm to perform the joint optimization of the RRH activation and robust CoBF by using the group sparse beamforming method through the weighted l1/l2 norm reformulation,where the group sparsity patterns of beamformers are used to guide the RRHs that can be switched off.Simulation results demonstrate that the proposed algorithm can significantly reduce the network power consumption by 28%in the low signal-to-interference-plus noise ratio scenario.In addition,the algorithm can approach the performance of the exhaustive search algorithm while having a much lower computational complexity.Finally,the fusion of massive MIMO and NOMA is investigated.The the-sis studies the gain of the downlink of the 5G with combined massive MIMO and the power domain NOMA technologies,and the gain of the power allo-cation scheme using the weighted maximum sum rate in actual channel case.Furthermore,we propose a joint quasi-orthogonal space-frequency coding(QO-STBC)and power allocation scheme in power,frequency and spatial domains,the proposed scheme has low complexity and outstanding performance.We can distinguish different users via reasonable power allocation,and distinguish the data layer belonging to a specific user by the diversity order.By QO-SFBC with joint multiuser coding,each data stream has the same detection diversity to ensure that each data stream has a similar performance.In addition,an optimization criterion for the coding rate of each layer of data stream under non-ideal interference cancellation is proposed,and the coding rate of each layer of each user stream is optimized according to the criterion.