Semi-Blind Channel Estimation Relying on Optimum Pilots Designed for Multi-Cell Large-Scale MIMO Systems

Channel estimation in the presence of pilot contamination-induced inter-cell interference (ICI) is a major challenge in large-scale multiple-input multiple-output (LS-MIMO) systems. In this paper, a subspace-based semi-blind channel estimator (SBCE) relying on an optimum pilot design (in the sense of minimizing the channel estimation error covariance) is proposed for LS-MIMO systems. The proposed SBCE is capable of exploiting the asymptotic orthogonality of the channel vectors encountered in LS-MIMO systems, while taking advantage of both the optimized pilots and the data symbols. In order to ensure the best-possible performance of the proposed SBCE, we analyze the properties to be satisfied by the optimal pilots and then design these pilots relying on Zadoff-Chu sequences. As a beneficial result, the intra-cell interference is completely eliminated, and the ICI is substantially reduced. Our analytical and numerical results confirm that the performance of the proposed SBCE is superior to that of the representative state-of-the-art channel estimators in practical LS-MIMO systems, which have a finite number of base station antennas and data symbols available to be capitalized on for channel estimation.