Robust Transceiver Optimization In Downlink Multi-Cell Multiuser Mimo Systems

With the development of wireless communication, wireless network is required to can accommodate more users, have greater system capacity and better user quality service, the goal of future wireless communication system is using the lowest frequency spectrum, power consumption and hardware complexity to provide a variety of high-quality high rate service. In the multi-cell MIMO downlink systems, interference has become a major obstacle of multi-antenna technology to improve system performance, a base station(BS) serves multiple users simultaneously, There will be intra-cell interference, and at the same time, the users will be subject to interference from adjacent cells, in addition, there is noise. How to optimize the transceiver system and design the transmitter precoding matrix and receiver has become one of challenges of communication system.Traditional methods usually assume that the CSI is perfect, in practice, however, perfect channel estimation is not achievable, the CSI that has error would significantly reduce the performance of the system, so traditional algorithms can’t be directly applied. What the paper discussed is in the multi-cell downlink MIMO system, when the channel estimation error is in a bounded region, for a series of optimization problems in which MSE and transmit power are treated as the performance indexes, how to obtain the optimal transceiver design through iterative algorithms and convex optimization toolkit.The first part introduces the background and significance of the task, outlines the research work in this thesis, points out the research direction, describes the robust optimization and convex optimization problems, and provides the theoretical basis for the later work. The second part discusses in the multi-cell downlink MIMO system, in the case of CSI error and the transmit power constrained, aiming at reducing the MSE of the user’s received signal, a robust transceiver scheme based on convex optimization theory is proposed. The errors in the channel state information are assumed to be bounded, therefore, the system’s MSE will also be in an uncertain region, by minimizing the worst-case MSE within this range, optimal transceiver design can be obtained, which thus can solve a range of MSE-optimization problems, including sum MSE minimization and max per-user MSE minimization. The Coordinated CB scene optimization can be achieved directly by utilizing the method of single-cell MIMO scene to extend, but in the JP scene, as each base station (BS) has individual power constraint, it is hard to optimize directly, this part takes minimizing the upper bound of the user’s received signal as the goal to optimize the design. Simulation results demonstrate that when there is CSI error, compared with the conventional schemes, the proposed scheme can enhance the quality of the system effectively and reduce MSE.In the third part, we change the point of view, in the case of CSI error and the MSE is constrained, study the robust transceiver design that sets the system transmit power as the goal and propose transceiver scheme based on convex optimization theory. Similarly, for the CB scene, the optimization design can achieved directly by utilizing the method of single-cell MIMO scene to extend, but for the JP scene, it is necessary to zoom the MSE function firstly, and then optimize the design, get the optimal transceiver scheme. The simulation results show that compared with other corresponding algorithms, the performance of the proposed algorithm is significantly improved.