Largest Generalized Eigenvector Precoder for CoMP-JT Massive MIMO Systems

Abstract

In this paper, we investigate the downlink (DL) precoder design in coordinated multi-point joint transmission massive multiple-input multiple-output systems. The DL precoder design problem is formulate to maximize the sum-rate under the per base station transmit power constraint. Utilizing the first-order condition, the structure of the optimal precoder is derived, involving the generalized eigenvectors of a pair of matrices. In accordance with this, the largest generalized eigenvector (LGEV) precoder is proposed to solve the first-order condition in an iterative manner, which involves solving the complicated generalized eigenvalue problem in each iteration. Specifically, we propose to solve the eigenvalue problem numerically for low-complexity implementation based on the inverse free Krylov subspace method. Simulation results demonstrate that the proposed LGEV precoder achieves satisfactory performances with fast convergences within a couple of iterations.