Research On User Scheduling And Joint User And Antenna Selection Of MU-MIMO System

In multiple-user MIMO system the same time-frequency resource could be shared by manyusers using space divine multiple address, which improves the spectral efficiency further butleads to the interference among users at the same time. To avoid the complexity of subscriberterminals, multiple-user interference has to be restrained at the base station throughpre-coding technique. Block diagonalization, one of the most effective pre-coding schemes, issuited for the situation where multiple antennas are deployed at each subscriber terminal.Multiple-user interference can be eliminated thoroughly by block diagonalization of lowcomplexity. However, there is a constraint of block diagonalization that the number oftransmit antenna should be no less than that of the sum of all subscriber terminals. A part ofthe receiving antennas have to be selected in order to satisfy the constraint.To make users with the implementation of block diagonalization pre-coding scheme inmulti-user multiple input multiple output system get fair service, a user scheduling algorithmbased on differential evolution was proposed. Differential evolution (DE) was first improvedso that the combinations of users of any size can be scheduled. Then, combine DE with theimproved proportional fair algorithm to schedule users. Simulations show that, the efficiencyof the proposed algorithm is improved substantially, and fairness as well as sum rate issuperior compared to the best existing algorithm for different SNR and different numbers ofusers, which can effectively solve the problem of user scheduling.Furthermore, to avoid bad performance of one specific antenna of the selected user anddecrease the complexity and costs of the hardware, an improved joint user and antennaselection algorithm is put forward. It is employed to search possible antenna subsets formaximizing system sum rate due to its fast convergence rate, high convergence accuracy andbrilliant global searching ability. Simulation results show that joint user and antenna selectionbased on differential evolution has better accuracy and, in most cases, higher efficiencycompared with the existing best algorithm.