Study On Key Techniques Of Closed Loop MIMO System With Imperfect Feedback

It is believed that the wireless MIMO technology will be one of the key ones that realize the high speed broadband wireless Internet access networks in the future and has wide application prospect in the third generation (3G) or the forth generation (4G) mobile communication. Relative to SISO, large gains in capacity of communication over wireless channels and significant improvement in quality of service for wireless communication are feasible in MIMO system with exploiting multi-path effect. Accordingly, the lasting demand for the communication and limited spectral resource are satisfied simultaneously.Closed loop MIMO system means that the information about channel state, which should be used to change the transmit scheme adaptively, can be feedback to the transmitter to improve the performance. Relative to the open loop system, it can increase the transmission speed and improve the reliability. The following factors such as time-varying of the wireless channel, realiability of feedback link and accuracy of the channel estimation, however, may influent the quality of channel state information at transmitter (CSIT). The dissertation is mainly about the study of technology in closed loop MIMO with imperfect CSIT for three system frames. The author obtains main results as follows:(1) Single-user MIMO systemA method for precoder design in the QOSTBC MIMO system is proposed under imperfect feedback. We consider the channel model with mean and transmitter correlation. With the consideration of feedback delay, an approximation method with minimum eignvalue is used to ensure the minimal precoding gain, which simplified the design to the situation under OSTBC. Simulation demonstrates the robustness and validity.(2) Single-user MIMO OFDM systemWith the goal of maximize transmission rate, a MIMO OFDM scheme with joint adaptive bit and power loading based on any-dimension beam-former is proposed under channel mean feedback. Combined with the full-rate and full-diversity of QOSTBC, any-d beam is used to increase the transmission rate. To reduce complexity of the adaptive designer, bi-section algorithm is used in adaptive bit and power loading. Compared to the OSTBC and Greedy algorithm, we show the performance gain with QOSTBC based on any-D beamform and bi-section method.A separated adaptive method for MIMO OFDM system under limited feedback is proposed. To reduce the computational complexity, adaptive beamform vector and adaptive bit loading are designed separately. A modified sphere interpolation algorithm is used to reduce the feedback bits for large number of subcarrier. We show the validity through simulation.(3) Multi-user MIMO SystemThe impact of imperfect feedback in multi-user MIMO system is analyzed. With quasi-static Rayleigh channel, estimation error and feedback delay, we derive the closed-form of outage capacity and average bit error rate in the TAS/MRC MU MIMO system. We demonstrate the impact of the imperfect factor, multi-user diversity and system antenna setting on the system through simulation.The above analysis is extened to the situation with Nakagami-m fading model. The analytical form of outage probability and average bit error rate are derived for the TAS/MRC scheme. The simulation demonstrates the impact of fading parameters, multi-user diversity and antenna setting.Considering fairness between users and quantity of feedback information, we propose a robust scheme under limited feedback in MU MIMO system. Through only permitting qualified users to feedback one bit information, the scheme can reduce the feedback bits effectively. At the same time, it can also against the feedback error. To ensure the fairness, a modified performance criterion is used to select qualified user and best transmit antenna. Estimation error is considered as compute the threshold. Simulation shows that the scheme can ensure the fairness under specific performance.