Channel Estimation for Multicell Multiuser Massive MIMO Uplink over Rician Fading Channels
Liang Wu,Zaichen Zhang,Jian Dang,Jiangzhou Wang,Huaping Liu,Yongpeng Wu
DOI: https://doi.org/10.1109/tvt.2017.2698833
IF: 6.8
2017-01-01
IEEE Transactions on Vehicular Technology
Abstract:Pilot contamination (PC) is a major problem in massive multiple-input multiple-output (MIMO) systems. This paper proposes a novel channel estimation scheme for such a system in Rician fading channels. First, the possible angle of arrivals (AOAs) of users served by a base station (BS) are derived by exploiting the channel statistical information, assuming a traditional pilot structure, where the pilots for the same-cell users are orthogonal but are identical for the same-indexed users from different cells. Although, with this pilot structure, the channel state information (CSI) derived contains CSI from other-cell users caused by PC, the line-of-sight (LOS) component of the desired user is PC-free when the number of antennas equipped at the BS is large. Then, based on the AOAs and the contaminated CSI, the LOS component of each user served by a BS is estimated, and data are detected by using the derived LOS components. Finally, the decoded data are used to update the CSI estimate via an iterative process. The achievable spectral efficiency of the proposed scheme is analyzed in detail, and simulation results are presented to compare the performance of the proposed scheme with that of three existing schemes.