| The multi-cell multi-user MIMO (MC-MU-MIMO) systems, which have attracted more and more attention, can provide a high spectral efficiency. In addition to the interference from other antennas, there are also interferences among users as well as base stations (BSs). Precoding techniques with power allocation schemes can be employed at BSs in order to suppress or eliminate the interference when channel state information is available at BSs. User scheduling can be employed when the total number of users is larger than that of users which BSs can be serviced at the same time. In view of the above issues or problems, this paper will focus on the following aspects:(1) In order to solve the problem of interference in multi-cell multi-user MIMO systems, and improve its sum-rate, the maximizing sum-rate precoding algorithm is proposed. By the simulation, we can obtain the following results:the proposed maximizing sum-rate precoding algorithm (MSR) and its improved version perform better than maximum signal-to-leakage-and-noise precoding algorithm (Max-SLNR), minimum mean square error (MMSE), and block diagonalization (BD), whereas the singular value decomposition (SVD) can achieve the optimal sum-rate performance. But, SVD requires full cooperation among BSs and among user terminals. This makes SVD difficult to implement in practical systems. However, it provides a upper bound of sum-rate for MC-MU-MIMO systems.(2) In order to further enhance the sum-rate performance of MC-MU-MIMO systems, Power allocation before precoding is investigated on the basis of (1). In order to consider the practical situation, we also investigate several power allocation schemes in a coordinated base station downlink transmission with total, per antenna and per base station power constraints. By simulation, we obtain the following results:the water-filling(WF) power allocation can further improve the performance of all precoding algorithms with total power constraints; Due to narrowing the range of the power allocation as power constrait varies from "total" to "per antenna", the performance of BD gradually become worse. For all power constraints, the modified water-filling (MWF) performs better than the WF in terms of sum rate.(3) To solve the problem of multi-user resource scheduling under the condition that the total number of users is larger than that of the users which BSs can be served at the same time, we not only study some classic scheduling schemes including TDMA-based maximum sum-rate scheduling, round-robin scheduling and proportional fairness scheduling, signal-to-Ieakage-and-noise (SLNR) based scheduling and signal to interference plus noise ratio (SINR) based user scheduling schemes. From simulations, we find the following results: the performances of the SLNR based a user scheduling schemes are the same for both lower bound and one-by-one, when the total number of users waiting for service is slightly bigger than that of users to be selected, it is noted that the former has lower complexity than the latter; when the total number of users is far bigger than that of users to be selected, the performances of SLNR scheduling scheme based on one-by-one is the best one among all compared scheduling schemes, its complexity is also lower and its fairness is almost the same as the SINR-based one-by-one scheduling. |
PDF Full Text Request