Research On Large Scale MIMO Transmission Technologies

The unique characteristics of mobility and personalized services make mobilecommunications adapt to the needs of the information age, and keep thriving vitalityand huge market potential. With the rapid growth and popularity of the mobile Internet,the data traffic of mobile communication networks is growing rapidly with annualgrowth rate doubled. The fact that the mobile communication services grow rapidly andthe wireless spectrum resources become more and more strained makes the large scaleMIMO technology be expected to address the issue of the lack of spectrum resources，tocope with the explosive growth of the demand on data transmission, and to meet theuser demand on data services. Large scale MIMO can support tens of even hundreds ofantennas, and make full use of the spatial degrees of freedom to service more userssimultaneously, resulting in improving the system capacity significantly withoutadditional bandwidth. Therefore, large scale MIMO undoubtedly becomes one of thekey technologies. In this context, we carry out in-depth study on of the large scaleMIMO transmission technology. The main contents are summarized as follows:1. On the basis of an analysis of the characteristics, performance advantages, andthe art of status, we discuss the problems including the channel capacity from theperspective of information theory and the minimum number of antennas, and evaluatethem by simulation. And then we design the large scale MIMO transmission, introducethe specific design and implementation of the overall transmit system and each functionmodule, discuss related technical solutions, and provide specific solutions to eachfunction module.2. A design approach of CSI reference signal (CSI-RS) applicable to32antennaport is proposed, simulated and evaluated. This CSI-RS with a hybrid of FDM andCDM can achieve a better trade-off between the performance and overhead. Besides, weanalyze the performance of different reference signals with different density in a largescale MIMO system.3. A hierarchical codebook design of multi-antenna quantized precoding foruniform linear array layout is proposed, simulated and evaluated. This codebook withthe Kerdock codebook as the main codebook and rotational matrices as thesubcodebook, can make up shortages of existing codebooks, and provide effective multi-antenna precoding operations, especially for channels with strong correlation.4. A double codebook design for uniform rectangular array layout is proposed,simulated and evaluated. This codebook can well portray channel characteristics, and isable to provide an effective solution to support multiple users as well as differentscenarios, even though the feedback overhead is quite limited.