Research On Spatial Division And Multiplexing In FDD Massive MIMO Systems

In Massive MIMO systems, the base stations are equipped with a large number (tens to hundreds) of antennas, simultaneously serving multiple terminals in the same time-frequency resources. Compared with MIMO technology, the massive MIMO technology can provide aggressive multiplexing gain, and enhance the spectrum efficiency and the system capacity. It has become one of the most important potential technologies for future mobile communication systems. This thesis focus on the downlink spatial division and multiplexing schemes of frequency division duplexing (FDD) massive MIMO systems. Detailedly, we research on the realization and optimization of the joint spatial division and multiplexing (JSDM) method and the application and extansion of spatial modulation (SM) and port modulation (PM) technologies.Firstly, the spatial division and multiplexing technology in multiple-antenna systems are summarized. The open-loop and closed-loop encoding schemes are studied and compared, including space division multiplexing and space diversity, as well as common precoding algorithms. Moreover, the methods of acquiring downlink channel state information (CSI) in massive MIMO systems are summarized especially in FDD massive MIMO systems.Secondly, the transmission schemes of JSDM under different scenarios are analyzed. The system model and precoding algorithm under one-ring model are described. After that, we provide an improved JSDM (IMP-JSDM) downlink transmission method which realizes the user grouping by adopting the beamforming processing at the receiver under the general scenario of SCM. Simulation results show that the JSDM transmission method achieves significant reduction of CSI feedback. Simulation results also show that the proposed IMP-JSDM method based on user grouping has good performance in FDD massive MIMO systems.Thirdly, we describe the system model and encoding mechanism of SM. The system performance and detection algorithms are discussed. The complexity of various detection algorithms are analyzed. Addtionally, a new angle-based ordered (ABO) detection algorithm for SM systems is proposed and simulated. Simulation results show that the proposed algorithm can obtain good bit error ratio (BER) performance and quite low computational complexity simultaneously.Finally, we research the application of the PM in single-user systems and multi-user systems. We propose a useful PM precoding method to realize the PM technology in multi-user massive MIMO systems. Simulation results show that the multi-user massive MIMO system adopting the proposed PM precoding method has better average BER (ABER) performance than that with the traditional multi-user precoding method, while significantly reducing the feedback for CSI estimation at base stations.