Research On Precoding Design And User Selection In Multi-antenna Systems

With the rapid development of wireless communications, the problem of higher demand of user rate and co-channel interference has become more and more challenging. MIMO has become the key technology of modern communication system due to its ability of achieving higher spectrum efficiency, and cognitive radio also is the key technology of future communication due to efficient reuse of the limited spectrum resources. However, how to efficiently user these key technologies, especially in the co-channel interference problem, still need further study.In this thesis we focus on research of linear precoding and multiuser selection. By design precoding scheme to suppress co-channel interference in multiantenna systems, and thus efficiently enhance spectrum efficiency. According to sources of interference, the precoding design is specified in multicell MIMO systems and cognitive radio MIMO system. For the multiuser MIMO system, this thesis fully explores the multiuser diversity by multiuser selection design, and thus further enhances the spectrum efficiency.For the multicell MIMO system, this thesis first proposes interference-aware uplink transmission schemes for two cell multiple-input multiple-output (MIMO) system. Unlike conventional transmission schemes without considering the interference probably caused to other cell, this thesis jointly optimize the transceiver beamforming vectors to maximize the desired signals while removing the intercell interference. Moreover when asymmetric interference is considered, this thesis gives a balancing beamforming (BBF) scheme where the interfering transmitter is to strike a compromise between maximizing the desired signal and minimizing the generated interference. For the general multicell MIMO system, this thesis proposes a distributed precoding scheme based on Max-SWGINR criteria to balance between maximizing the desired signal power and minimizing the generated interference. The proposed schemes get good tradeoff of the complexity and performance. Simulation results also verify the superiority of the proposed schemes.For the interference problem in cognitive MIMO system, this thesis carried out full study of precoding design based on different mode of spectrum sharing. Different from the ideal CSI assumption, this thesis first propose a cognitive frame structure, in which a sensing phase is exploited to estimate the effective interference channel information. For the opportunistic spectrum sharing cognitive mode, this thesis proposes an enhanced SLNR-based precoding scheme to maximize the secondary sum rate while guaranteeing zero-interference at the primary receiver. For the performance limited problem of primary and secondary users, this thesis gives a cognitive relay based two-slot cooperative spectrum sharing strategy, the slot time allocation and beamforming vectors are given. Simulation results validate the superiority of the proosed schemes. Both cognitive and primary users get large enhancement due to the cooperation.At last, for the multiuser MIMO system, this thesis gives user selection schemes to fully explore the gain of multiuser diversity. For the problem of complexity, the thesis first proposes two low-complexity user scheduling algorithms for single-cell multiuser system. Then for the user selection problem in cognitive MIMO system, this thesis proposes a double-parameter user selection algorithm based on the orthogonality between primary and secondary channels. This thesis also gives the complexity analysis and asymptotic performance of the proposed algorithms. Simulation results verify the superiority of the proposed algorithms.This thesis studies the co-channel interference problem of multicell MIMO system and cognitive MIMO system based on different mode of spectrum sharing, and user selection problem in multiuser MIMO system, then proposes efficient precoding schemes and user selection algorithms to fully enhance the system performance and explore gain of multiuser diversity. These studies improve the performance of the system and have good usability.