Research And Performance Evaluation On Massive MIMO Pre-coding Technology

Massive multiple-input multiple-output (Massive MIMO) technology is one of the key technologies for the future 5G mobile communication system. In this system, the base station is deployed with the antenna array which is formed by dozens or even hundreds of antennas to send data to multiple users, simultaneously. The pre-coding technology is used to improve the performance in terms of the system capacity and spectrum efficiency. However, the system transmits a lot of data streams at the same time and frequency so that the Massive MIMO system is a serious self-interference system. Therefore,in order the improve the reliability of the system,some signal processing techniques must be used to suppress interference. Moreover, in consideration of the size of the antenna array used by the system and the signal processing capability limitations of the user krminals,the way that the receiving end use the detection technology is not feasile. Therefore, the pre-coding technology has become the main interference suppression technology in AeMassiveMIMOdownlink.Based on this background,the thesis focuses on the non-linear pre-coding technology for the multiuser Massive MIMO downlink system. The main contributions of this thesis are as follow.Firstly, the thesis analyzes the linear pre-encoding and non- linear pre-encoding.For example, the Imear pre-encoding based on the zero-forcing crkerion and the minimum mean square error criterion as well as the vector perturbation non- linear prc-encoding.Secondly, this thesis devises a vector pre-coding scheme on the condition that the transmitting end has the imperfect channel state information. In this scenario, this thesis assumes that the transmitting end can obtain the second-order statistics information of the channel estimation error matrix to establish the imperfect channel information model. In this model, this thesis uses the channel matrix geometric mean decomposition for channel estimation matrix,and then deduces the pre-coding matrix and the power control factor under the zero-forcing criterion and the minimum mean square error criterion,respectively. And,this thesis proposes a low complexity descent method by iteration for searching the perturbation vector in the complex field through improving the existing perturbation vector search algorithm. Thereout,this thesis achieves a robust vector pre-coding scheme under the condition of the transmitter knowing the imperfect channel state information. Under the uncorrelated Gaussian channel condition, the simulations show that this scheme showed a good bit error rate performance and convergence. And, as the number of transmit antennas and the number of users increases, this pre-coding scheme can obtain a better bit error rate performance.Finally, combining with LTE downlink physical layer key technologies, such as CRC checking, code block segmentation, Turbo encoding and decoding, rate matching, code block concatenation, scrambling, modulation and demodulation and MIMO-OFDM, A LTE link layer simulation platform is built by this thesis. The baseband signal processing procedure is realized. Simulation results show that in the same environment and simulation parameters, the proposed pre-coding scheme present a lower block error rate than the pre-coding scheme based codebook, and achieve improved system performance.