Lattice Reduction Aided Transceiver Design for Multiuser MIMO Downlink Transmissions

In this paper, we present a linear joint transceiver design for downlink transmissions of multiuser multi-input multioutput (MU-MIMO) system using lattice reduction algorithms, based on the minimum sum of the mean-squared errors (MSEs) criterion and subject to a per-user power constraint. Our design consists of two main steps. First, we decompose the MU-MIMO system into parallel single user (SU-) MIMO systems using existing multiuser interference elimination methods (e.g., Block diagonalization, generalized zero forcing channel inversion), and then for each user, we formulate a transceiver optimization problem in lattice-reduced domain with the goal of minimizing sum of the MSEs of information symbols. The problem is solved sub-optimally with the help of lattice reduction algorithm for each user. We compare the error performance and complexity of our proposed design with several existing schemes through simulations and show that our new design performs very well with low complexity especially when each user is equipped with a large number of antennas (e.g., More than 3) and receives spatially multiplexed data streams.