Low-rank matrix fitting based channel estimation for full-dimensional massive MIMO systems

Full-dimensional (FD) massive MIMO can overcome the limitation of the physical space at the base station (BS) and meet the increasing capacity requirement for future wireless communications. However, the pilot overhead for the acquisition of the channel state information (CSI) at the BS could be prohibitively high due to the large number of antennas. In this letter, exploiting the Kronecker property of the channel, we propose a low-rank matrix fitting based channel estimation (LMFCE) scheme to reduce the pilot overhead in FD massive MIMO systems. The simulation results illustrate that the proposed LMFCE scheme can significantly reduce the pilot overhead for the channel estimation.