Research On Beamforming Technique In Large-scale Antenna System With Limited-feedback Channel Information

The large-scale antenna technology can effectively improve the energy efficiency and spectrum efficiency of the communication system without increasing the transmission power and the bandwidth of the system, by simultaneously serving dozens of users through the multiuser beamforming technique. The multiuser beamforming technique needs to acquire the user's instantaneous channel information to calculate the beamforming vector of each user. In the frequency-duplexing large-scale antenna system, the acquisition of this information requires a high pilot overhead, and the feedback overhead will also be great. Fortunately, we can reduce the dependence on channel information by exploiting the spatial correlation of large-scale antenna arrays, but the performance of this beamforming technique is limited. Therefore, this paper mainly studies the beamforming technology in large-scale antenna system with limited feedback of channel information.In the second chapter, we analyze and simulate the relationship between the large-scale antenna array and its spatial correlation, and explain the reason why the spatial correlation can be used in the large-scale antenna system. However,the performance of this beamforming technique is limited and it can not make full use of the advantages of large-scale antenna technology.In the third chapter, we improve the beamforming technique based on spatial correlation, proposing a new beamforming technique based on spatial correlation and signal leakage. In this technique, we select part of the antenna to send pilot symbols, then the receiver get the information of the antenna channel estimation and feedback them to the base station which uses these information to calculate a conversion matrix. With this conversion matrix, we transform the beamforming matrix based on space correlation, so as to achieve the purpose of improving system performance. In this process, we have a certain amount of pilot symbols and information feedback cost in exchange for system performance improvement, the greater the cost, the better the performance.Based on the simulation results with ideal feedback, the performance gain of the novel beamforming technique is about 3dB, 4dB and 6dB respectively when the antenna selection ratio is 25%, 33% and 50%.In chapter 4, we analyze the shortcomings of the new beamforming technique proposed in Chapter 3 under the condition of limited feedback. It has low deployment flexibility and high computational complexity. In order to improve the deployment flexibility, we remove the relation between the process of constructing the beamforming weight vector and the selected set of antenna indices, and then replace the original calculation algorithm with algorithm with similar performance and low computational complexity. According to the simulation results under limited feedback, the performance gains of the improved schemes are 11.88%, 19.40% and 21.55%, respectively, when the number of feedback bits is 6,7,8, compared with the spatial correlation.In summary, the beamforming technique proposed in this paper can reduce the dependence of the instantaneous channel state information by using the partial channel state information to improve the spatial-correlation-based beamforming technique, so as to reduce the frequency division duplex large-scale antenna System pilot symbols and feedback overhead.