Research On Precoding In MIMO System

Improving the spectral efficiency by multi-input multi-output (MIMO) technology without increasing the system bandwidth and antennas transmitting power, it is the common view on the design of communication system standards in future. Precoding is one of the core technologies to guarantee the transmission performance of MIMO. The correlative algorithms of precoding affect the validity and reliability of the system directly. The paper regarded precoding technology as the entry point, and the single user as well as multi-user precoding technology in MIMO system was investigated. The main contents of the paper are as follows:1. The paper did the research on single user precoding based on the different types of CSI. A joint transceiver design algorithm in LOS (line of sight) correlated channels with imperfect CSI was provided. The algorithm introduced the idea of joint design, and the percoding&decoding matrices were solved by iteration. The algorithm deduced a channel estimation error model based on training sequence, which made the characterization of channel estimation error more practically. The influences of substream numbers, LOS component, correlation of the channel and channel estimation error on the performance of the algorithm were investigated by simulation. And also, the paper studied the limited feedback precoding. The codebook of precoding and power allocation together based on an existing precoding structure was designed; the codebook design used VQ method, and gained the satisfied BER performance when the feedbacks were low.2. The paper studied the precoding in multi-user MIMO system. Considering of the interference cancellation precoding, the paper analyzed the feasibility and validity of substream selection for block diagonalization (BD) algorithm and provided the substream selection algorithm based on BD; the substream selection improved the BER performance greatly with a bearable capacity loss. In the next, the paper introduced the idea of BD into joint design, provided a joint transceiver design algorithm in LOS correlated channels with imperfect CSI for multi-user MIMO system (BD+MMSE). The algorithm made the users'channels independently by block diagonalization, and then the users'procoding and decoding matrices could be obtained by iteration respectively. The proposed algorithm had the advantage of BD which processed the users independently and had a better BER performance at the same time. Considering of the interference suppression precoding, the paper provided an improved transceiver design with iterative initial value optimization and Lagrange factor linearization based on the minimum total mean square error criterion for the downlink MIMO. Compared to the existing method, the proposed took full advantages of the channel state information and had a faster convergence speed. On the other hand, the proposed algorithm could change the progress of obtaining the Lagrange factor in a linear way instead of the original nonlinear means, got the analytic solution of Lagrange factor and made the factor more accurately. The proposed algorithm improved the BER performance of the system efficiently and reduced the computational complexity distinctly at the same time.3. The paper did the research on power allocation and antenna selection based on multi-user precoding. Considering of the fairness and QoS demands among the users, the paper provided a SNR control power allocation scheme. The scheme could make all the users have the same BER performance. Also could control SNR of the users flexibly. Then the paper did a research on unified power allocation for interference suppression precoding, provided an iterative waterfilling algorithm for maximizing the sum capacity, especially in correlated channels. Regarding the interference as noise, the algorithm introduced the idea of classical waterfilling to the interference system. The simulation results indicated that the algorithm proposed improved the sum capacity efficiently in correlated channels. In the study of antenna selection, the paper combined antenna selection with multi-user precoding, and improved the sum capacity of the system by two classical antenna selection algorithms.