Research On Mimo Precoding Techniques Based On Spatial Multiplexing

MIMO Pre-coding can get spatial diversity gain and multiplexing gain through configured both the transmitter and receiver by multiple antennas, the spatial diversity gain and multiplexing gain allowing the system to achieve higher channel capacity, spectral efficiency and lower system bit error rate. However, MIMO causes interference between multiple parallel signal streams and multiple users, and because of the limited processing capability and power control, it is not practical to recover data only relying on the MIMO detection techniques at the receivers, so MIMO pre-coding technique is proposed to solve such problems. MIMO pre-coding pre-processing the data to be transmitted at the transmitter based on the channel state information, effectively suppressing interference signals in parallel, not only can significantly improve the system capacity, but also the complexity of the process can be simplified to a great extent the receive side.This thesis mainly discusses the existing pre-coding techniques, and analyzed the key factors that affect MIMO pre-coding performance. Firstly, the linear pre-coding scheme based codebook including spatial multiplexing and transmit diversity pre-coding techniques are introduced in this thesis. Both the strategy of codebook design algorithm of calculation the pre-coding matrix indicator and algorithm of MIMO detection at the receiver are also analyzed. Secondly, the linear pre-coding techniques which are not based on codebook including ZF (Zero-Forcing) criteria and MMSE (Minimum-Mean-Square-Error) are researched in detail. there are three algorithms based on the channel matrix decomposition scheme:SVD (Singular Value Decomposition, SVD) decomposition, GMD (Geometric Mean Decomposition, GMD) decomposition and UCD (Uniform Channel Decomposition, UCD) decomposition pre-coding techniques. Taking into the complexity of all these linear pre-coding algorithms, a pre-coding algorithm based on LDR matrix decomposition is analyzed in detail by this thesis. Not only the complexity of the new algorithm is much smaller than that of other decomposition scheme, but also the performance is better than SVD scheme. Thirdly, the non-linear pre-coding techniques including Costa pre-coding, TH pre-coding and GMD-TH pre-coding algorithm are also studied. The simulation results shown that the performance of THP scheme based on Cholesky decomposition is better than that of the THP one based on QR decomposition, but worse litter than that of the THP one based on GMD.Finally, a simulation platform which is in strict accordance with the3GPP LTE standard is build. All of the simulation results which is including the modulation order, MIMO spatial correlation and channel types are operation by this platform, and also give a special attention to the bit error rate performances between the perfect channel state and not perfect CSI conditions.