| Mobile Internet and the Internet of Things (IoT) will become the main forces driving the evolution of mobile communications towards 2020 and beyond. In order to meet the challenges of Mobile Internet and IoT, 5G arises at the historic moment. As one of the key technologies of 5G, Full-Dimension MIMO (FD-MIMO) utilizes the information of both horizontal and vertical dimensions to distinguish users and can effectively reduce the inter-user interference. At the same time, it improves the antenna gain. Therefore, it can provide higher capacity for the communication system and meet people's growing requirements for communication capacity. On the RAN # 65, 3GPP setted up the project of FD-MIMO. The key research contents are CSI-RS measurement, CSI feedback and so on. In low speed scenarios, the channel estimation precision based on the reference signal are affected by both inter-cell and intra-cell CSI-RS interference; In high speed scenarios, UE moves too fast and the channel state changes very quickly. And CSI feedback may not accurately reflect the UE's current channel state. It will result in the mismatch of precoding strategy which will bring more interference to other users and also lower the QoS of the served user. Therefore, research on interference suppression of the reference signal in low and high speed scenarios is of great importance for improving system performance. The main research contents can be summarized as follows:For the research on interference suppression of reference signal with different cells configuring the same CSI-RS resources, this thesis proposes a dynamic port selection scheme based on beamformed CSI-RS.In this scheme, UE measures the usage information of reference signal in adjacent cells by ZP CSI-RS and feedbacks it to the served cell. The served cell determines the assignment of CSI-RS ports for UEs with game theory based on the usage information of reference signal in adjacent cells. Simulation results show that the proposed scheme can solve the interference problem of reference signal with different cells configuring the same CSI-RS under different CSI feedback periods. The performance gain of the proposed scheme will be more obvious when more users require services in one cell.For the research on interference suppression of reference signal with different cells configuring difference CSI-RS resources, this thesis proposes a dynamic switch of resource utilization scheme based on beamformed CSI-RS. In this scheme, BaseStation (BS) collects the usage information of CSI-RS in adjacent cells, and decides if it should use the unoccupied CSI-RS resources of adjacent cells based on the collected information. It can make a good trade-off between CSI-RS interference suppression and resource utilization improvement. Simulation results show that the proposed scheme can achieve a 10% gain in cell-edge user throughput and a 8.4% gain in average user throughput compared with the CSI-RS nulling scheme.For the research on outdated PMI in high speed scenarios, this thesis proposes a dynamic PMI adjustment scheme. In this scheme, UE will trigger the base station to send aperiodic beamformed CSI-RS when its QoS is below a certain level. Then UE will measure the aperiodic beamformed CSI-RS and feedback the beam selection information to BS.The BS will adjust the outdated PMI based on the feedback information.Simulation results show that the proposed scheme can solve the problem of outdated PMI in high speed scenarios under different speeds and different CSI feedback delays. The performance gain of the proposed scheme will be more obvious when the speed is higher. The proposed scheme is suitable for solving the problem of outdated PMI in high speed scenarios.In conclusion, this thesis studies the interference suppression of FD-MIMO from the above three aspects. The interference suppression schemes of the reference signal in low and high speeds scenarios are mainly studied. Three patents have been applied. The performances of the proposed schemes are also simulated. |
PDF Full Text Request