Research On Adaptive Transmission For MIMO Heterogeneous Broadcast

To provide high speed and high quality multimedia services in multi-user system, multi-antenna technology, which can provide the high frequency spectrum efficiency, has been introduced for the new generation wireless communication system. Through transmit beamforming technologies at transmitters and receive combining technologies at receivers, multi-antenna technology can exploit full spatial multiplexing diversity and multi-user diversity, then greatly enhance the system throughput. However, most antenna technologies are based on two assumptions: the perfect channel state information at the transmitter and all the users'channels obey the independent identical Rayliegh distribution. However, all these assumption are not applicable in the real systems because the users in the same system may not be homogeneous and the uplink feedback channel may be speed restricted. Limited feedback strategies and transmit strategies based on partial channel state information in heterogeneous multi-user system become hot topics recently. In this dissertation, we mainly focus on heterogeneous multi-user multi-antenna broadcasting channels with imperfect channel state information. In-depth researches has been conducted on how to accurately capture and highly utilize the efficient channel state information in different heterogeneous multi-user scenarios.First, this paper thoroughly analyzes the multi-user scenarios with heterogeneous channel distributions. In this situation, traditional limited feedback strategies at the user ends can lead to a serious waste on system feedback overhead. To solve this problem, an adaptive multi-user feedback strategy based on channel statistic information is proposed. The proposed method can effectively trace the change of the channel distributions, adaptively change the user's feedback code-books, accurately aim at the exact channel directions and significantly reduce the feedback overhead based on long-term channel statistic information under different channel environments. Here we also conduct researches on multi-user schedule scheme when different users are experiencing different channel distributions. The proposed group based proportional fairness scheduler can achieve very good tradeoff on maximize the system throughput and maintain fairness among all users.In this paper, we also analyze the multi-user scenario with heterogeneous channel fading parameters, which means users'channel obey the same distributions but have different parameters due to users'different situations. For a scenario of which users'rank order of channel matrix is different, we propose an adaptive multi-user multi-stream transmit scheme based on limited feedback. This scheme effectively decreases the overhead of downlink feedback signaling as the user number increases. By adaptively transmitting single stream or multiple streams based on different rank of users'channel matrices, this scheme can greatly enhance the throughput for every single user. Furthermore, this scheme can eliminate the interference among the sub-streams for each user, which, in a result, improve the system throughput performance.