Downlink Adaptive Transmission In FD-MIMO Distributed Antenna Systems Exploiting Partial Channel State Information

As one of the key technologies of future mobile communication system,massive multiple-input multipleoutput(MIMO)technology can make full use of space resources,reduce transmission power,improve system spectrum efficiency and energy efficiency.In order to further explore the space resources as well as deploying more antenna elements,full-dimension MIMO(FD-MIMO)system with two-dimensional(2D)uniform planar array(UPA)at the base station(BS)has attracted considerable attentions.However,in frequency-division duplexing(FDD)mode,the acquisition of channel state information(CSI)at the BS becomes a bottleneck as the number of antenna grows large.Downlink transmission algorithm exploiting only statistical CSI becomes an effective way to reduce the cost of CSI feedback.Moreover,compared with the centralized antenna system(CAS)which deploys a single base station in a cell,the distributed antenna system(DAS)can effectively enhance the coverage,reduce the link interruption,and enhance the system capacity by adding the distributed remote radio unit(RRU)in the system.The combination of DAS and FD-MIMO can make good use of their advantages to enhance the system performance.Therefore,this thesis studies the downlink adaptive transmission strategy using statistical CSI for the distributed FD-MIMO system.Firstly,a power allocation strategy which exploits only statistical CSI for distributed FD-MIMO downlink transmission system is proposed.For scenario where both the BS and the RRU are equipped with twodimensional UPA array,a downlink 3D beamforming algorithm and a user scheduling algorithm are proposed based on maximizing the lower bound of the user's received average signal-to-leakage-and-noise-ratio(SLNR).Then,the approximate expression of ergodic sum rate of the system is derived,and a downlink power allocation algorithm maximizing the ergodic sum rate approximation,which relates only to the statistical CSI of each user,is obtained.Simulation results show that the approximate expressions of the ergodic sum rate are very close to the Monte Carlo simulation results.The proposed user scheduling algorithm and power allocation algorithm can significantly improve the system performance.Secondly,a cooperative transmission strategy based on frequency reuse is proposed for the distributed FD-MIMO downlink transmission system.First,all users in the cell are divided into two categories: central serving users and jointly serving users,according to the large-scale factors between each user and RRU.Among them,each RRU's central serving users constitute a central serving user set of this RRU,and all the jointly serving users in the system constitute a jointly serving user set.The users in each RRU's central serving user set are served only by their corresponding RRU.And the users in the jointly serving user set are served collaboratively by multiple RRUs.Each RRU's central serving user set shares the same frequency resource,and the jointly serving user set uses different frequency resources,so as to eliminate the interference between them.Two types of transmission scheme are proposed for the jointly serving user set: full cooperative transmission and partial cooperative transmission.In the fully cooperative transmission scheme,all RRUs transmit to the scheduled users in the jointly serving user set.In the partial cooperative transmission scheme,user in the jointly serving user set is transmitted by one or more RRUs with acceptable transmission performance.Then,the approximate expressions of ergodic sum rate of jointly serving user set and central serving user set are derived under the two proposed cooperative transmission schemes.Simulation results show that the proposed scheme can improve the transmission performance of users in the jointly serving user set,so that the users which are far away from the RRUs can get high transmission performance.And the approximate expressions of the ergodic sum rate are very close to the Monte Carlo simulation results.Finally,based on the above cooperative DAS downlink transmission system,a dual-layer precoding and grouping transmission method for the central serving user sets is studied,and then extended to the jointly serving user set.First,for the central serving user sets,a user scheduling algorithm maximizing the sum rate is proposed that exploits only statistical CSI.For the scheduled users,a low complexity user grouping method using only each user's statistical CSI is proposed,and a dual-layer precoding transmission scheme is designed.In this transmission strategy,users in the same user group use the same outer precoding matrix,which only depends on the users' statistical CSI.The inner precoding vectors of these users depend on the lower-dimensional effective CSI of the users in each group.Finally,the proposed transmission strategy is extended to the jointly serving users under the full cooperative transmission and partial cooperative transmission respectively.Simulation results show that the proposed scheme can achieve high ergodic sum rate with small amounts of CSI,and the proposed user grouping algorithm can flexibly achieve the tradeoff between the amount of CSI required and system performance by choosing the appropriate number of user group.Compared with the full cooperative transmission scheme,the partial cooperative transmission scheme of the jointly serving user set can further reduce the amount of CSI required and improve the transmission performance.