LR-RZF Pre-coding for Massive MIMO Systems Based on Truncated Polynomial Expansion

In order to effectively eliminate multi-user interference at the transmitter, the transmit signals are needed to be pre-processing, which is called pre-coding. Traditional linear pre-coding algorithms, especially regularized zero forcing (RZF) pre-coding, are famous for good performance and low computation complexity. However, they cause noise amplification which requires high transmit power to prevent. Lattice-reduction aided (LR-aided) technique is used to deal with the row/column of matrix, which can make the matrix orthogonality better. Therefore, to avoid noise amplification as well as effectively eliminate the multi-user interference, we propose a LR-RZF pre-coding algorithm which based on matrix truncated polynomial expansion (TPE) with J terms. TPE method can decrease the complexity of matrix inversion in RZF. Compared with RZF pre-coding, LR-RZF pre-coding has lower bit error rate.