Research And Design On Channel Feedback Scheme Of Massive MIMO For 5G System

Massive multiple input multiple output( Massive MIMO) technology, which is the key technology of the next-generation wireless communication, can improve the capacity and robustness of wireless communication system. The remarkable advantage of Massive MIMO depends on whether the transmitter can achieve the channel state information. However, with the increase of the number of antennas in Massive MIMO system, the amount of feedback increases dramatically. Besides, the feedback channel capacity is limited, so it faces great challenges to feedback the channel state information to transmitter. Therefore,the study of channel feedback technology of Massive MIMO system has great significance.In this thesis, the channel feedback schemes of massive MIMO system are studied and designed to new application scenarios. Firstly,the compressive sensing based and dimension reduction based channel feedback scheme is studied. The large number of antennas arranged in massive MIMO lead to strong spatial correlation between antennas. Thus, by taking advantage of this property, the feedback information can be compressed by using compressive sensing and dimension reduction methods. The implementations of the two feedback schemes are studied and the performance of the two schemes are analyzed and evaluated by simulation. Secondly, considering the temporal correlation between the channel information of the adjacent time slots in massive MIMO system, a channel spatial correlation and time correlation based channel feedback technology in massive MIMO system is studied in this thesis, and a differential compressive channel feedback scheme for massive MIMO is proposed. The performance of this feedback scheme is analyzed and evaluated. Simulation result shows that the differential based feedback scheme can perform better than traditional compressive sensing based feedback scheme. Finally, this thesis proposes the compression quantization based massive MIMO channel feedback scheme based on the compressive sensing technology and vector quantizition technology. By taking advantage of the channel structure features,the channel information is compressed and quantized by using compressive sensing method and improved codebook at the terminal. Simulation result shows that the proposed feedback method can decrease the feedback bits and improved the feedback efficiency.