Research On Mu-Mimo Precod-Ing For Cooperating System

With the wider spread of mobile handheld terminals, the require-ments of mobile wireless data service are increasing from users, which continuously demands for better data rate and spectral efficiency. It has been proved that Gbit is achieved from the breakthrough of technology in rapidly evolving LTE-A, which is the renewal of LTE. Therefore, a series of the specifications for IMT-Advance are ensured by LTE-A.Multiple Input Multiple Output (MIMO) are enhanced in LTE-A, which enables maximal8-layer precoding to realize spatial division mul-tiplexing (SDM). The urban microcell are chosen as a practical study in this dissertation based on LTE-A downlink MU-MIMO, which makes it possible for separating transmission of multiple data streams. The suffi-cient conditions have been derived in this dissertation based on the re-search of Nash Equilibrium, which is the balance for user requirements. Besides, the numerical simulation has been taken for optimal channel ca-pacity. The main innovations of this dissertation are shown below:1) The sufficient conditions of Nash Equilibrium are derived based on the Foschini-Miljanic (FM) algorithm, providing achievable SINR (signal to interference and noise ratio) for each user. In limited feedback system, the channel state information at the transmitter (CSIT) is imper-fect, which causes the co-channel interference (CCI) among all users. Obviously, it is inevitable for users to compete for their required SINRs. It has been proved that with users iteratively resetting their power level to the desired ones, exponentially fast convergence to the desired perfor-mance occurs finally in any system with CCI. The sufficient conditions for the FM algorithm are derived in this dissertation based on above re- search, and many math tools like Theory of Matrices are used in the pro-cess. Not only the achievable SINRs are guaranteed, but also the effective measures to improve SINRs are shown in these sufficient conditions Under the condition of RVQ (Random Vector Quantization), it is more efficient to decrease the CCI caused by quantization error rather than simply increasing the transmit power.2) The comparative evaluation of Power Control (PC) schemes. Supposed perfect CSIT, Waterfilling turns out to be the optimal PC scheme to maximize the overall capacity. However, the requirement that the transmitter has accurate CSI is difficult to meet, as well as getting the analytical expression of the optimal PC scheme mathematically. Most preceding researches of precoding capacity simply used equal power set-ting (EP) in limited feedback system, which will cause a waste of power and a decrease of QoS (Quality of Service) in practical systems. In this dissertation, PC factors are combined with different precoding methods, and the specific received signals are derived for those different schemes. It can be seen that PSO schemes are superior than EP ones by capacity gain up to10.8%, whose numerical simulation combined MIMO, limited feedback, precoding, PSO, D2D and some other efficient technologies.3) The research on the cooperative scenario with D2D. As a novel scheme, D2D enriches the structure of conventional cellular system, and it has many advantages to improve data rate required by users. Conven-tional cellular system and D2D are combined in this dissertation, and throughput of different PC schemes are analyzed in this system. The spectrum reuse of D2D under the control of cellular base station will cause interference to other cellular users. So the spectrum allocated for D2D should be remote to cellular users. Precoding could be employed for the spectrum reuse without the constraint of distance mentioned above, which is more useful for smaller cell in the future.