Multiuser MIMO systems in single-cell and multi-cell wireless communication
MetadataShow full item record
MIMO technology improves the capacity and link robustness of wireless communication by deploying multiple transmit and receive antennas. A multiuser MIMO communication system involves multiple mobile stations (MS) and potentially multiple base transceiver stations (BTS). These systems are fundamentally limited by interference, and require new treatment of both the capacity characteristics and physical layer algorithm design. In this dissertation, multiuser MIMO systems in both single-cell and multi-cell environments are studied. A single-cell MIMO broadcast channel is defined by a central BTS transmitting to multiple MSs simultaneously over the same spectrum. A multi-cell MIMO system consists of multiple BTSs transmitting to MSs in different cells. For a single-cell MIMO broadcast channel, block diagonalization is a transmit precoding technique that multiplexes multiple users in the spatial domain and pre-cancels inter-user interference. Precoder can be adaptively designed based on the size of transmit/receive antenna arrays and the number of users. In the case where the BTS has more antennas or radio frequency (RF) units than strictly required for interference cancellation, this dissertation proposes novel downlink precoder with enhanced transmit selection diversity. Eigenmode selection and transmit antenna selection are derived to optimize a symbol error rate upper bound and improve the diversity performance. When there are a large number of users in the system, a subset of users and receive antennas may be selected to maximize the sum capacity under the block diagonalization signaling. The optimum joint user and antenna selection involves brute-force search, therefore is prohibitively complicated. In this dissertation, two low-complexity sub-optimal selection algorithms are proposed to significantly reduce the selection complexity. Conventional single-user MIMO techniques suffer significant performance loss in an interference-limited multi-cell network. Interference on a MIMO system is more severe than in a single-antenna cellular network, as each antenna element acts as a unique interferer. In this dissertation, power control is investigated as an interference management tool to properly determine the transmit power of MIMO array under a pre-determined SNR constraint. Two uplink MIMO power control techniques are proposed. The first equal allocation algorithm enforces each antenna element of a MIMO array to transmit at the same power, resulting in a closed-form but suboptimal solution. The second algorithm adaptively distributes power on a MIMO antenna array to exploit the channel selectivity, hence substantially reduces the transmit power and interference, and creates far better cell coverage. Finally, block diagonalization precoding in the single-cell scenario is generalized to the multi-cell environment as a coordinated MIMO transmission technique. Multiple BTSs cooperate with each other to design the downlink signal, thereby eliminating interference and improving the spectral efficiency. An improved precoding scheme is proposed to address the per base station power constraint in the cellular environment. Future research topics for cellular block diagonalization precoding are discussed.