Research On Multi-user Selection In Multi-antenna Multi-user Systems

Multi-antenna technology is a key technology for fourth-generation wireless system, which could provide spatial multiplexing and spatial diversity for wireless communication system and greatly improve the spectrum efficiency and quality of wireless communication system. The nature of independent time varying channels across different users in a multi-user wireless system provides multi-user diversity. By exploiting the multiuser diversity, multiuser systems offers higher throughput than the single user systems, while allowing transmitting different services to different users simultaneously. User selection is key problem for exploiting multiuser diversity to improve system performance. In this paper, we focus on the research of the user selection problem. Especially for user selection problem and corresponding feedback strategy based on the special space division multiplexing access(SDMA) technology, provide several transmission schemes, which have good tradeoff of the overhead, computational complexity and throughput.The thesis investigate nonlinear Tomlinson-Harashima precoding(THP) and the corresponding user selection method in-depth. THP have better power efficient. Based on the specificity of THP, we proposes a low complexity greedy user selection algorithm based on heuristic Gram-Schmidt orthogonalization for the downlink multi-antenna multi-user systems with THP. The proposed algorithm can exploit the multi-user diversity, improve the throughput performance. The performance of the proposed algorithm is very close to the throughput performance of the optimal selection algorithm. THP with HGS-Greedy scheme have better throughput performance than the ZFBF-SUS scheme and gradually approaches the capacity of DPC as the number of users increases, which demonstrates this scheme has the asymptotic optimality in the capacity domain. Moreover, the complexity of the algorithm is low, which is suitable for practical systems.The thesis investigate orthogonal SDMA precoding and the corresponding user selection method and limited feedback strategy in-depth. Precoding with orthogonal precoding vector is an promising scheme for the practical systems. The Per-User Unitary Rate Control(PU2RC) scheme has been accepted by the 3GPP-LTE standard. In practical systems, the number of the active users is moderate or low in a single cell. In this system, we investigate the a novel orthogonal SDMA precoding, which has better throughput performance than the ZF-SDMA in the moderate SNR section with perfect CSI. In practical systems, it is difficult to obtain the perfect CSI and the research show that it is no need to obtain it. In the FDD model system, the transmitters get the CSI through the limited feedback. Based the novel orthogonal SDMA precoding, we propse a novel user selection method and feedback strategy in the systems with small number of users. The expression of the feedback CQI is derived and the feedback strategy and user selection method is provided. The proposed scheme has better performance than the scheme of limited feedback and user selection base on ZF-SDMA, provides better reference for practical systems design.