Downlink Mu-mimo Precoding And User Scheduling Techniques

MIMO (Multiple Input Multiple Output) technology has a wide application in multi-user system, for achieving higher date rates and providing much reliable wireless communication links. Taking advantage of Space Division Multiplexing Access (SDMA) technology, MU-MIMO systems can communicate with multiple users at the same time slot and frequency, which fruitfully improves the efficiency of frequency spectrum. However, in the downlink of MU-MIMO system, users can not cooperate with each other. Thus, they can't do the interference cancellation through the multi-user joint detection technology as in the uplink. So the technique of suppressing multi-user interference (MUI) through preprocessing in the downlink MU-MIMO system becomes a hot topic in MU-MIMO systems. In MU-MIMO system, on the one hand, the base station does the precoding for the user signals before transmission, which greatly reduce or cancel the MUI. On the other hand, the base station will select the users whose channels are in good condition through certain optimal scheduling schemes for pairing and transmission, which can make the best of multi-user diversity gain. This dissertation mainly focuses on the precoding technologies and scheduling schemes in MU-MIMO systems. In the following, we will give a detailed description about our research work.Firstly, we will have a concrete description about the precoding technology in the downlink of MU-MIMO system. The current existing Block Diagonalization method has the disadvantage of higher complexity. Triggered by this, we propose a new BD precoding algorithm, which can greatly lower the computational complexity without any system capacity loss. It is to find the null space of interference matrix in another way, which is based on the QR decomposition and the inverse of triangular matrix.Secondly, we will have a analysis about the user scheduling algorithm in the downlink of MU-MIMO system. The SUS (Semi-orthogonal User Scheduling) algorithm can be improved by MRC (Maximum Ratio Combining) to overcome the shortage that no diversity gains of receiving antennas can be obtained. However, it is also unable to solve the problem that the SUS algorithm can't make use of the multiplexing gain of receiving antennas. Thus, in this dissertation, we propose a totally new user scheduling algorithm based on the capacity lower bound, which can make the best of diversity gain and multiplexing gain of receiving antennas.At last, the joint precoding and user scheduling algorithms under limited feedback will be analyzed concretely. In the dissertation, a joint algorithm based on the chord distance is proposed. The quantization error and the multiuser interference are estimated by the metric of chord distance. Compared with the conventional joint algorithm based on the zero-forcing channel inversion, the proposed joint algorithm can both reduce the user feedback and adverse impact caused by quantization error.