Research Of User And Antenna Selection Technique For Multiuser MIMO Systems

There has been considerable recent interest in wireless multiple input multiple output (MIMO) communications systems, due to their potential for dramatic gains in channel capacity. More recently, attention has shifted to multiuser MIMO systems. For multiuser MIMO systems, antenna selection and user selection are worth studying specially, because antenna selection gets MIMO system gain with low complexity and lowcost, user selection enhances the system capacity through selecting part most superior users, effectively. This paper studies several techniques in the multiuser MIMO systems, including antenna selection, user selection. The main contents of the paper are listed as follows:1. For the multiuser MIMO downlink with block diagonalization precoding, two suboptimal user and antenna selection algorithms which are based on the channel Frobenius norm are proposed in the following two cases. In the first case : each user is allocated certain number of antennas , a antenna for a user at each step is selected to maximum F-norm of the projection for system; and in the second case : the total number of antenna of all users is allocated , a antenna at each step is selected to maximum F-norm of the projection for system, and two algorithms have less computational complexity than the corresponding optimal algorithms.2. Considering user selection and power allocation for multiuser MIMO downlink. Each user employs a zero-forcing receiver to detect signal, and feeds back the post-detection SNR of each stream to the base station, and BS allocate power of each antenna with water-filling to enhance system capacity. we use 2-D water-filling along time and frequency theorem to allocate the transmit power, and the results show that 2-D water-filling has the higher rate than 1-D water-filling, particularly in low SNR situation.3. For multiuser MIMO systems based on space-time block code, we analyze the performances of two user selection schemes. The first scheme is that the BS selects only one user with the largest SNR at a time slot while the second scheme is that the BS selects multiple users with the largest SNRs at a time slot. Furthermore, these two schemes are compared by theory analysis and simulation. Simulation results indicate that the first scheme exhibits significant performance gains over the second scheme in the term of capacity, but the second scheme can serve multiple users simultaneously.