Abstract
Low complexity vector-projection-aided vector perturbation precoding is proposed. Vector perturbation is a well-known precoding approach for near-capacity multiantenna multiuser communication. The optimal perturbation vectors are obtained using the sphere encoder. A suboptimal lattice reduction algorithm (Babai's approximation) is applied to reduce the complexity. In this thesis, we propose to use vector projection technique to find the perturbation vector. Furthermore, to obtain a much better perturbation vector, we propose two-stage vector perturbation scheme which perturbs the data twice with a little more complexity. Moreover, we try to perturb the signal vector instead of perturbing the data vector in use of idle antennas via vector projection. Simulation results show that two-stage vector perturbation is slightly better. Besides, vector-projection-aided vector perturbation can apply to a reduced precoding scheme (lattice reduction aided vector precoding) which is not suitable for the use of Babai' approximation. Simulation results show that our proposed method is a great substitute for sphere encoding.