A Low-Complexity Linear Precoding Algorithm Based on Jacobi Method for Massive MIMO Systems

In massive multiple-input multiple-output (MIMO) systems with the increase of the number of received antennas at base station (BS), linear precoding, as zero-forcing (ZF), is able to achieve near-optimal performance and capacity- approaching due to the asymptotically orthogonal channel property, but it involves matrix inversion with high computational complexity. To avoid the matrix inversion, in this paper, we propose a novel low-complexity linear precoding algorithm based on Jacobi method (JM). The proposed JM-based precoding can achieve the near- optimal performance and capacity-approaching of the ZF precoding in an iterative way, which can reduce the complexity by about one order of magnitude. Furthermore, the convergence rate achieved by JM-based precoding is quantified, which reveals that JM-based precoding converges faster with the increasing number of BS antennas. Simulation results show that JM-based precoding achieves the near-optimal performance and capacity- approaching of ZF precoding with a reduced number of iterations.